UBE2D2 · ENST00000398733.8

EXTENDED 178 aa (canonical 147 aa) · UniProt P62837 · CDLMPS

chr5:139561698:+:ATC:ENST00000398733.8

AI summary N-terminal extension folds confidently alone but sits structurally disconnected from the UBC catalytic core, with no domain, localization, or targeting change.

How it diverges

The added 32-aa N-terminus is highly disordered and proline/arginine-rich (whole-protein disorder shift, S2) yet also folds with high per-residue pLDDT in isolation — but PAE between the extension and the shared UBC body is very high (~18 Å) with only a handful of contacts confined to the junction, meaning it does not integrate as a real structural addition to the catalytic domain. No InterPro domain is gained or lost, DeepLoc/SignalP/TargetP calls are unchanged, and the shared UBC-like fold (which starts downstream of the extension) is essentially untouched (RMSD 0.23 Å at high pTM).

Why it matters

UBE2D2's known function depends entirely on its compact UBC catalytic core (E1 charging, backside Ub engagement, RING/HECT E3 cooperation) and its cytosolic/endosomal localization — none of which this extension's mechanism findings disturb: the core fold, domain content, and predicted compartment are all preserved. The extension reads as a disordered, poorly-integrated N-terminal appendage rather than a functional module, so on the mechanism evidence alone it has no clear bearing on UBE2D2's established E2-conjugase role.

Biophysical shift

LLM confidence low

No mass-spec peptide validation of the extension, and the P1 'structured' pLDDT reading is undercut by very high inter-region PAE and minimal contacts, so the biological reality of this extension as a folded or functional appendage remains uncertain.

Folding

Canonical (147 aa)

Download CIF

Isoform (178 aa)

Download CIF

Coloured by ESMFold2 pLDDT confidence (blue = high, orange/red = low). Differential region — residues 1–31 (added in isoform) — recoloured on a yellow→purple pLDDT ramp so it stands out. Drag to rotate · scroll to zoom · download a CIF to explore in your own viewer.

Canonical PAE

Isoform PAE

Predicted aligned error: expected Cα error (Å) at residue j when the fold is superposed on residue i. Dark = confident relative placement; bright = uncertain. The dashed outline marks the differential region (1–31).

Evidence — click any tile for the differential-region detail

C Conservation Interesting

LLM reasoning

The N-terminal extension this isoform adds ahead of the canonical UBE2D2 start is itself under real purifying selection, not just passively retained sequence: unique-region phyloP is 3.82 (well above the ~2 strong-constraint threshold), and amino-acid identity across primates is 96.8% and across mammals 82.3%, both clearing their respective thresholds. As expected for a shorter span, frac_intact (87% primates, 44% mammals) is not compared to the canonical span's own frac_intact since that metric is confounded by length. The mean-identity comparison is the informative one here, and while it trails the near-invariant canonical mean identity (99.8% in both primates and mammals), a nearly 32-residue extension holding 82-97% identity deep into mammals argues this added segment is a genuine, conserved coding feature rather than incidental upstream sequence picked up by alternative translation initiation.

Unique region 96.8% similar across primates

Conservation metric	Canonical ORF	Differential ORF
Mean AA identity	100%	97%
Frame intact (fraction of species)	88%	87%
Species aligned	25	23
Species frame-intact	22	20
Start codon conserved	96%	96%
Deepest intact species	Microcebus_murinus	Microcebus_murinus
Phylo depth (MRCA)	7	7

Unique region 82.3% similar across mammals

Conservation metric	Canonical ORF	Differential ORF
Mean AA identity	100%	82%
Frame intact (fraction of species)	75%	44%
Species aligned	20	18
Species frame-intact	15	8
Start codon conserved	94%	71%
Deepest intact species	Loxodonta_africana	Canis_lupus_familiaris
Phylo depth (MRCA)	12	11

Unique region PhyloP: 3.82 — purifying selection

Track	Differential	Shared	vs shared
phyloP mean	3.82	5.77	0.663
phastCons mean	0.838	0.976	—

Property	Canonical	Isoform
Start codon	ATG	ATC
Kozak context (−9..+4)	CCTCCCACCATGG	CCGACCGAGATCG
phyloP at start codon	6.79	1.83
phastCons at start codon	1	0.03
phyloP over Kozak window	4.77	1.36
phastCons over Kozak window	0.858	0.162
Kozak mismatch — full consensus	3	6
Kozak window GC content	0.692	0.692

D Detection Interesting

LLM reasoning

This extended N-terminal TIS is robustly detected at the transcriptional/translational level: reproducible across 5 of 6 cell lines with highly significant p-values (e.g. HeLa p=1.6e-9, K562 p=2.3e-16), and its initiation efficiency is substantial and consistent (0.019-0.172 across lines), reaching about 0.45x the canonical start's usage rather than being a rare or noise-level event. Peptide-level proteomic confirmation is absent, however - all 7 isoform-unique peptides spanning the added N-terminal sequence were detected but none passed PepQuery2 validation, so direct protein-level evidence for translation of this extension is currently lacking. On balance the strong, multi-line ribosome-profiling/TIS signal argues this alternative start site is genuinely and efficiently used at the RNA/translation stage, even though mass-spec confirmation of the resulting protein has not yet been achieved.

detected in 5/6 cell lines

Cell line	Canonical	Isoform	p-value
HeLa	27.5	12.4	1.56e-09
K562	9.88	8.07	2.3e-16
U2OS	2.8	2.14	1.09e-08
RPE1 Async	2.96	2.03	6.63e-07
RPE1 Que	1.71	0.947	2.74e-06

alt used 0.45× vs canonical

Cell line	Canonical	Isoform
HeLa	0.382	0.172
K562	0.113	0.0926
U2OS	0.0429	0.0328
RPE1 Async	0.0371	0.0254
RPE1 Que	0.0343	0.019

0/7 isoform-unique peptides validated

Feature	Canonical	Isoform
Tryptic peptides (in-silico)	10	7
Validated by mass-spec	0	0
Isoform-unique peptides	—	7

L Localization Neutral

LLM reasoning

The N-terminal extension shows no meaningful predicted localization change: DeepLoc top-compartment call is unchanged (Cytoplasm|Nucleus for both isoform and canonical), with nearly identical top probabilities (0.656 vs 0.691) and identical nuclear localization/export signal calls. The only flagged difference is a membrane-association shift from 'Peripheral|Soluble' in canonical to 'Soluble' in the isoform, but this is a minor drop of a secondary peripheral-membrane annotation rather than a compartment reassignment, and probabilities across all compartments (extracellular, cell membrane, mitochondrion, ER, Golgi, lysosome, peroxisome) are all broadly similar between the two. No secretory signal peptide or mitochondrial/chloroplast transit peptide is gained or lost — both remain SignalP 'OTHER' and TargetP 'noTP' with negligible probability deltas (~0.0006-0.0016). Overall the added 32-aa N-terminal segment does not redirect the protein to a new subcellular destination or introduce/remove a targeting signal.

iso: Cytoplasm, Nucleus | canon: Cytoplasm, Nucleus

Property	Canonical	Isoform
Predicted location	Cytoplasm\|Nucleus	Cytoplasm\|Nucleus
Sorting signals	Nuclear localization signal\|Nuclear export signal	Nuclear localization signal\|Nuclear export signal
Membrane	Peripheral\|Soluble	Soluble

iso: noTP | canon: noTP

Predictor	Canonical	Isoform
TargetP	noTP	noTP
SignalP	OTHER	OTHER

M Mutation Landscape Not interesting

LLM reasoning

No meaningful disease signal anywhere on this isoform, and the germline-constraint inputs are structurally uninformative here since this is an extension (the added 32-aa N-terminal segment was 5'UTR/intronic, never coding). Disease-variant density argues against, not for, functional relevance: only 3 disease variants fall in the unique region versus 58 in the shared core (enrichment ratio 0.25), and none of the 4 total ClinVar entries on this isoform are pathogenic — all are Uncertain significance, and all 4 sit in the shared canonical region (isoform positions 67, 153, 169, 177), none near the alternative start codon. The histogram shows just 4 distinct positions with a single hit each — too sparse to call a cluster. COSMIC contributes only 3 hits in the unique region, each with cosmic_sample_count=1 and annotated intronic, i.e. negligible recurrence. Germline variants in the unique region (136 gnomAD hits) are consistent with never-coding sequence (intronic consequence, ESM-C out-of-distribution, 0 constrained positions) and cannot be read as tolerance or constraint of a protein region.

gnomAD variants 5.92× more in unique region — tolerant

Variant set	Differential	Shared	Depletion ratio
gnomAD variants	136	109	5.9×

Property	Differential	Shared	Enrichment
ESM-C mean LLR	-1.92	-0.0173	—
Constrained positions	0	0	—

Variant set	Differential	Shared	Enrichment
Scorable variants	131	109	5.7×
Damaging variants	30	31	4.6×
— of which loss-of-function	30	2	71×
AlphaMissense-pathogenic	0	21	0×

Property	Differential	Shared
Mean ΔLLR (ESM-C)	-0.681	-3.36
Min ΔLLR (ESM-C)	-2.97	-11.4
Mean AlphaMissense	—	0.56

Disease variants 4.08× less in unique region — depleted

Variant set	Differential	Shared	Enrichment
Disease variants	3	58	0.25×
Pathogenic	0	0	—

Variant set	Differential	Shared	Enrichment
Scorable variants	3	58	0.25×
Damaging variants	1	41	0.12×
— of which loss-of-function	1	3	1.6×
AlphaMissense-pathogenic	0	34	0×

Property	Differential	Shared
Mean ΔLLR (ESM-C)	-0.621	-7.34
Min ΔLLR (ESM-C)	-1.24	-11.9
Mean AlphaMissense	—	0.787

P Predicted Structure Not interesting

LLM reasoning

The N-terminal extension is confidently folded on its own (per-residue pLDDT 0.80-0.98, mean 0.908) but this reads as a locally well-formed segment with no resolved relationship to the rest of the protein: mean PAE between the extension and the downstream body is 17.8 A (up to 31 A), and contacts are limited to only 10 total, concentrated at the immediate junction (residues 33-36) rather than a genuine surface patch, with essentially nothing beyond it. No secondary-structure element in the extension qualifies (longest is a 5-residue strand at 0.86 pLDDT, below the 6 aa length threshold), so there is no gained helix/strand to speak of. The shared core is unperturbed and this is well-supported, not an artifact: RMSD 0.23 A against high global pTM (0.87-0.88) and high shared-region pLDDT (0.885-0.971). Together this is a confidently-folded-but-unintegrated appendage with no secondary structure and an unchanged core — nothing here argues for functional consequence.

pLDDT Differential Region: 0.909

Metric	Canonical	Isoform
pLDDT (whole protein)	0.885	0.961

Metric	Differential	Shared	Enrichment
pLDDT	0.909	0.823	1.1

Shared-Region RMSD: 0.23 Å

Property	Canonical	Isoform
Shared-region pLDDT	0.885	0.971

No helix or strand in diff region

Metric	Differential	Shared
Alpha helices	0	4
Beta strands	0	2
Longest element (aa)	0	15
Mean pLDDT	—	0.97

Isoform-unique	Shared core
—	alpha helix 32–46 15 aa · pLDDT 0.96
—	beta strand 63–68 6 aa · pLDDT 0.96
—	beta strand 79–87 9 aa · pLDDT 0.97
—	alpha helix 130–142 13 aa · pLDDT 0.98
—	alpha helix 152–159 8 aa · pLDDT 0.97
—	alpha helix 162–174 13 aa · pLDDT 0.97

Isoform-unique	Shared core
beta strand 2–6 5 aa · pLDDT 0.86	beta strand 52–55 4 aa · pLDDT 0.98
—	beta strand 97–100 4 aa · pLDDT 0.98

S Structural Characteristics Neutral

LLM reasoning

The added 32-aa N-terminal segment does not overlap any real InterPro domain (the UBC-like fold hits all sit downstream in the shared canonical region), so the extension neither disrupts nor gains a defined domain. Whole-protein biophysical shift did fire on disorder fraction, driven by a highly disordered, low-complexity, proline/arginine-rich unique region (disorder fraction 0.385 unique vs 0.082 shared, low-complexity fraction 0.90 unique vs 0.0 shared) against an otherwise small whole-protein disorder delta (0.053) and negligible charge delta (-0.002), suggesting the extension is an appended disordered tail rather than a change to the folded core. The sparse-autoencoder magnitude check did not clear threshold (top shared-feature shift 6.04 vs required 10.0), so despite large gained/lost feature counts (153/34), there is no strong evidence of a qualitative interpretability shift beyond what the biophysical disorder signal already captures. Combined, this points to a disordered N-terminal appendage with no clear domain-level or strong compositional consequence — a mild, unremarkable structural signature rather than a compelling functional signal.

No diverging domains

Feature	Canonical	Isoform
InterPro domains	9	10
Short linear motifs	3	3

less hydrophobic (-0.68) · similar charge · more disordered (+0.30)

Property	Differential	Shared	Enrichment
Isoelectric point (pI)	12.7	7.73	1.64
Hydropathy (GRAVY)	-1.03	-0.348	2.96
Fraction charged	0.226	0.238	0.948
Disorder fraction	0.385	0.0821	4.69
Disorder-promoting	0.839	0.517	1.62
Low-complexity fraction	0.903	0	—
Prion-like fraction	0.226	0.225	1.01
LLPS score	0.327	0.13	2.52
π–π propensity	0.194	0.265	0.729
Aromaticity	0	0.102	0
Instability index	104	42.9	2.43
Shannon entropy	2.93	4.19	0.698
Normalized complexity	0.677	0.969	0.698

187 SAE features differ

Feature	Label	Description	Activation	Prevalence
#448	N-terminal leader/targeting segments	N-terminal leader/targeting segments: the feature activates on the extreme N-terminus (first 2–5 residues) of proteins, especially the N-region of signal peptides and organelle transit peptides, and more generally on short, disordered/low-complexity N-terminal tails; strongest at residue ~3 and largely independent of amino-acid identity, occurring across all taxa and functions.	16.61	2
#4992	Small-residue disordered N-terminus	Context-dependent free N-terminus signature: short, disordered amino-terminal stretches beginning with the initiator Met followed by residues compatible with MAP cleavage or N-acetylation (e.g., M–G/A/S/T/V, and other contexts including M–D/E/L); strongest emphasis on small residues at positions 2–3; activation is absent when the native N-terminus is structured or starts with a signal peptide/transmembrane segment.	15.46	5
#7996	Generic extreme N-terminus detector	Generic extreme N-terminus detector: marks residues around position 5 (occasionally 6–7) at the very start of the polypeptide, most often within the N-region of signal/transit peptides of secreted or organelle-targeted precursors, but also in generally disordered N-termini of cytosolic/nuclear proteins; not motif- or residue-specific	10.80	2
#3609	Position-2 N-terminus residue detector	Detector of the identity of the second residue at the extreme N-terminus of proteins, with strongest preference for small residues (especially Ser, Ala, and Gly; also Thr/Pro), i.e., an N-terminal position-2 signal commonly associated with generic, disordered starts and co-translational N-terminus processing	10.78	2
#11504	Initiator methionine at M1	Initiator methionine at the very start of the polypeptide chain (M1), i.e., the translation start residue, independent of protein family, taxonomy, membrane association, or presence of signal/leader/propeptide regions; often annotated as post-translationally removed and typically situated in a flexible, coil-like N-terminus.	10.62	2
#5241	Early N-terminal positional signal	Generic early N-terminus positional signal peaking at residue ~5–7, found broadly within the disordered N-tail of proteins, including both cleavable leader/targeting segments (signal peptides, propeptides) of precursor proteins and disordered N-terminal regions of mature proteins; residue identity is not specific	9.59	3
#7986	Intrinsically disordered low-complexity regions	Compositionally biased, intrinsically disordered low‑complexity regions used as flexible interaction/nucleic‑acid–binding modules—enriched in Pro/Ser/Lys/Glu/Gln/Thr/Gly/Arg and encompassing poly‑Pro tracts, Lys/Arg‑rich basic stretches, and acidic Asp/Glu clusters—common in transcription/RNA‑processing factors, viral basic regulators, and the C‑terminal scaffolds of RNase E	6.68	22
#7026	Catalytic domain helix/loop boundary motif	A localized internal-domain motif that marks a specific short helix/loop boundary within a catalytic domain, situated downstream of the catalytic proton-donor residue in α/β hydrolase-like folds.	6.38	3
#4180	Hinge residues; C-terminal linker motifs	Secondary-structure transition/hinge residues—particularly helix termini and adjacent coil/turn positions—and conserved C-terminal/linker sequence motifs in nucleotide-binding domains	6.32	2
#6178	Nuclear basic/polar low-complexity IDRs	Intrinsically disordered, low‑complexity regulatory segments of nuclear proteins—especially transcription factors/cofactors, chromatin regulators, RNA/splicing factors, and nuclear RING‑type E3 ligases—enriched in basic (Lys/Arg) and polar (Ser/Pro/Gln/Glu/Gly/Thr) tracts that serve as activation/repression, partner‑binding, or localization (NLS‑like) regions; typically outside folded DNA‑binding/catalytic domains and often overlapping phosphorylation sites or simple helices within IDRs.	6.05	7
#6484	Prion-like low-complexity IDRs	Polar low-complexity intrinsically disordered regions (IDRs)—especially Q/N-rich (prion-like) segments and S/T- and glycine-rich tracts with simple SG/TG/PG repeats—typically located in inter-domain linkers and terminal tails of eukaryotic proteins and often harboring Ser/Thr phosphorylation sites.	5.53	6
#11753	Cytosolic N-terminal regulatory tails	Cytosolic N-terminal tails and disordered/low-complexity regulatory segments at the extreme N-terminus of eukaryotic signaling proteins, including phosphatases (classical PTPs, PP2C family) and membrane receptors/ion channels. Sequence character is mixed: some tails are Ser/Gly-rich or basic/acidic low-complexity stretches, others are short hydrophobic/charged N-terminal helices preceding a catalytic domain. The feature avoids transmembrane cores, catalytic/active-site loops, metal-binding residues, and well-folded globular interiors.	5.04	34
#5482	N-terminal targeting presequences	Universal eukaryotic N-terminal targeting presequences: the feature detects short, cleavable leader regions at the extreme N-terminus that direct proteins to organelles or the secretory pathway—especially chloroplast/apicoplast transit peptides and thylakoid lumen signals, but also mitochondrial targeting peptides and classical signal peptides. These segments are Ser/Thr- and small/hydrophobic–rich, enriched in Lys/Arg and depleted of acidic residues, typically low-structure/low-confidence and ending at the maturation cleavage site.	4.85	2
#3398	Ser/Pro-rich disordered tails	Serine/proline–rich low-complexity intrinsically disordered segments, especially terminal tails, linkers, and N‑terminal signal-peptide regions of small or secreted proteins; structured catalytic cores are largely ignored.	4.82	5
#4012	Proline-directed IDR phosphorylation	Intrinsically disordered Ser/Thr phosphorylation hotspots, with a strong preference for proline‑directed motifs (S/T‑P) characteristic of CDK/MAPK-like kinase targets across diverse eukaryotic regulators and viral phosphoproteins.	4.77	3
#9145	Helix–loop boundary/linker signal	A generic helix–loop boundary/linker signal: the feature prefers short coil segments and helix-capping/hinge positions that connect or terminate alpha-helices, especially interhelical loops in multi-pass membrane proteins and structured active-site/cofactor-binding loops in soluble enzymes; it tolerates diverse amino acids but often includes small helix-modulating or interfacial aromatic residues.	4.73	2
#12620	Disordered low-complexity regions	Intrinsically disordered, low‑complexity, compositionally biased regions/tails (IDRs), typically enriched in Ser/Gly/Pro/Ala/Thr and often occurring as acidic (Asp/Glu) or basic (Arg/Lys) tracts or Gln/Asn‑/Gln‑rich repeats; these segments are common in secreted precursors, viral proteins, micropeptides, and testis‑associated proteins, and can also occur as low‑complexity termini or surface loops appended to otherwise folded enzymes; they frequently coincide with low predicted structural confidence.	4.67	9
#12818	N-terminal targeting and disorder	N-terminal targeting and processing segments of secreted/endomembrane and organelle-targeted proteins—cleavable signal peptides, N-terminal signal‑anchor helices, and mitochondrial/chloroplast transit peptides—and the immediately downstream low‑complexity/disordered propeptide or luminal/matrix-facing stem; more generally, a preference for intrinsically disordered N‑terminal tails (including Ser/Thr/Ala‑rich plant segments), with occasional hits on short flexible N-termini of soluble cytosolic proteins.	4.66	29
#2530	LTV1/MPP10 motifs, TruD helix	Sequence-specific motif feature active in ribosome/snoRNP biogenesis factors and TruD-family pseudouridine synthases. In LTV1 homologs the feature recognizes a conserved internal "SSVxRRNEQL"-like motif, and in MPP10 homologs it recognizes a conserved "P(A/V)PVITEE" motif. In TruD/PUS7-family proteins it activates along an internal alpha-helix within the TRUD catalytic domain.	4.66	3
#7338	Amphipathic coiled-coil helices	Coiled-coil–like amphipathic alpha-helices with heptad-repeat character (hydrophobic a/d layers often L/A and charged/polar e/g positions enriched in E/Q/K/R/S/T), as found in long dimeric/oligomeric scaffolds, adaptors, and motor tails; the signal also extends to shorter amphipathic helices in peptide precursors.	4.51	10
#2038	Disordered low-complexity regulatory regions	Low-complexity, intrinsically disordered regulatory regions enriched for serine/threonine and glutamine/asparagine (often with glycine/proline/histidine runs), typically located in N- or C-terminal tails and flexible linkers of eukaryotic proteins; these segments include transcriptional activation/repression regions and phosphorylation-modulated interaction sites, while structured catalytic or DNA/RNA-binding domains are not targeted.	4.43	19
#11732	GH18/GH13 catalytic core loops	Conserved sites within the catalytic core of GH18 chitinase-like proteins and GH13 α-glucosidase/maltase-family enzymes, with peaks typically falling on loop/helix regions internal to the catalytic domain rather than on terminal tails.	4.43	3
#13964	Short amphipathic alpha-helices	Short amphipathic alpha-helical segments enriched in leucine (and other hydrophobes) with interspersed basic residues, most prominently the helices that compose HEAT/ARM-like alpha-solenoid repeats but also analogous isolated N-terminal helices in diverse proteins	4.38	3
#9395	Short terminal targeting segment	Terminal or short internal targeting/interaction segment: a short, compositionally biased or hydrophobic patch near a protein terminus, or occasionally an internal low‑complexity loop, that functions as a signal peptide or signal‑anchor (Sec/ER/thylakoid/Tat entry) when hydrophobic, or as a basic/Ser/Lys/Arg/Pro‑rich low‑complexity segment that mediates localization or partner binding in soluble proteins; typically the single prominent terminal segment of short secreted, membrane, or regulatory proteins	4.35	4
#5021	Polybasic intrinsically disordered regions	Low-complexity, often intrinsically disordered regions in short proteins, precursors, and microproteins across taxa, including viral accessory proteins, neuropeptide/hormone precursors, sperm nuclear proteins, flexible enzyme tails/inserts, and antisense-derived/uncharacterized microproteins. Activation is biased toward, but not restricted to, basic (Lys/Arg) clusters, with frequent peaks also at Pro/Ser/Gly residues in disordered context.	4.11	7
#8461	Ciliary/meiotic charged coiled-coils	Signal for residues within long α-helical coiled-coil segments of cilia/flagella- and meiosis-associated structural proteins, firing on heptad-repeat positions in charged, Glu/Lys/Arg-rich stretches.	4.08	3
#12751	Ser/Pro-biased low complexity regions	Compositionally biased regions (IDRs or short low-complexity segments) that are explicitly Ser/Pro-biased or enriched for simple dipeptide repeats (SS/PP, RS/SR, RG/RGG) and basic residues; these tracts occur in viral accessory proteins and micropeptides and as Ser/Pro-rich linkers/termini in diverse proteins. Generic hydrophobic signal peptides without such polarity/repeat bias are not targets.	4.06	4
#9962	S/T/P-rich disordered regulatory tails	Intrinsically disordered, low‑complexity regions enriched in serine, threonine, proline and polar/charged residues—flexible regulatory linkers/tails and propeptide segments that often host short linear motifs (e.g., phosphorylation- and proline‑rich motifs) and proteolytic processing sites; signal is absent from well‑folded catalytic domains and is common across taxa	4.05	5
#11023	Low-complexity disordered regions	Intrinsically disordered, low‑complexity segments—often N‑terminal tails or leader/signal regions—enriched in Ser/Pro/Gly/Arg and predicted as coils/low‑confidence structure; the feature marks flexible, poorly structured regions rather than a specific function	3.97	4
#11676	Disordered Ser/Thr phosphorylation sites	Short linear motifs centered on serine/threonine within intrinsically disordered regions that correspond to eukaryotic Ser/Thr phosphorylation sites (often SP/TP or basic R/K-flanked motifs), used for regulatory control across diverse proteins including viral proteins	3.97	4

Feature	Label	Description	Activation	Prevalence
#6362	Kinase domain helix-capping residues	Residues at helix–coil transition (helix-capping/linker) positions within the conserved protein kinase catalytic domain, especially in the C-lobe around long helices and the activation-segment vicinity; a generic structural marker of kinase cores rather than active-site motifs, conserved across Ser/Thr, Tyr, and dual-specificity kinases (including eukaryote-like bacterial kinases)	7.42	2
#7265	N-terminal initiator Met recognition	Recognition of the absolute protein N-terminus—specifically the initiator methionine (M1; often N-formylmethionine in bacteria/mitochondria) and the immediate +1 residue that becomes the mature N-terminus upon N-terminal methionine excision	5.41	2
#735	Aromatic edge beta-strand motifs	Short edge beta-strand segments in beta-sheet–rich domains (especially jelly-roll/beta‑sandwich folds such as Ig-like, cupredoxin, velvet and LPMO), typically at terminal/edge strands or loop–strand junctions, often containing or flanked by aromatic residues (Y/F) and frequently adjacent to a cysteine that participates in a disulfide or metal‑ligand; common in extracellular/secreted proteins but not restricted to them.	4.49	2
#8428	Noncatalytic beta-strand and loop hotspots	Residue-level signature within compact beta-rich domains, firing at sparse positions within or immediately adjacent to beta-strands and in short connecting loops, generally non-catalytic and common across diverse folds.	3.77	2
#10932	Globular hydrophobic core packing	Hydrophobic packing within well-folded, soluble α/β domains—i.e., core-facing residues in helices and strands of globular regulatory/signaling domains (kinase catalytic cores, bacterial HTH/sigma DNA‑binding domains, etc.); a generic structural/fold signal rather than a motif- or active‑site–specific feature.	3.21	2
#1785	RlmK/L dual-motif peaks	Short conserved sequence motifs within bacterial ribosomal RNA large-subunit methyltransferases (RlmK/RlmL family), with peaks on hydrophobic/aromatic residues embedded in two characteristic stretches of the enzyme.	2.70	2
#11070	Aromatic-polar loop micro-motif	Short polar/aromatic micro-motif centered on H or N typically followed by an aromatic (F/Y) and embedded in coils, turns, or short beta-strands; often located in active-site or ligand-binding loops, prominently adjacent to catalytic/binding residues in N-acetylhexosaminidases (GH20) and chitinases, but reused across diverse enzyme folds and outer-membrane transporters.	2.65	2
#5114	Short N-terminal targeting arm	N-terminal segments (typically the first ~10 residues) across a wide range of bacterial, viral, and eukaryotic proteins. These are often signal/targeting/assembly arms, including RNA/DNA-binding tails, secretion/chaperone-binding regions, and assembly linkers.	2.32	5
#6919	Helix caps and turns	Residues at alpha-helix boundaries and short interhelical turns (helix caps/hinges), including the first/last few positions of soluble or transmembrane helices and the brief loops linking paired helices in helical repeat units; enriched for small residues (especially Gly/Ala, also Asn/Thr/Val/Pro) consistent with helix-capping and turn formation	2.21	2
#8838	P450 heme-binding Cys pocket	Cytochrome P450 heme-binding Cys-pocket and its upstream “meander” loop—i.e., the conserved C‑terminal heme‑ligand region including and immediately preceding the FGxGxxxCxG signature that carries the axial heme‑coordinating cysteine; a fold-level element shared across diverse P450 families and taxa	2.14	2
#15589	Mid-chain hydrophobic core	Mid‑chain, well‑packed structural core segment of small domains—a contiguous, hydrophobic (Gly/Val/Leu/Ile‑rich) secondary‑structure run that stabilizes the domain (most often a β‑strand block or β‑α‑β unit; in all‑α folds the analogous central helix pack; in membrane proteins a mid‑bundle TM helix), typically adjacent to but not itself the catalytic/ligand‑binding residues.	2.13	3
#12567	N-terminal amphipathic leader helix	A short, compositionally biased N-terminal segment around positions ~18–40 that is enriched in charged/polar residues (often with glycine runs) and behaves as an intrinsically disordered patch with strong propensity to form a single short amphipathic alpha-helix; this generic “leader-like” interface is used across proteins as a targeting/secretion/assembly or interaction module (rather than a specific sequence motif).	2.09	9
#10622	Amphipathic helix assembly motifs	Short amphipathic α‑helical interface segments (8–20 aa) enriched in Leu/Ile/Val and acidic residues with interspersed Lys/Arg and nearby aromatics, used as helix–helix packing/assembly or protein–protein interaction motifs. The feature recurs in helical scaffolds of retroviral/retrotransposon Gag/capsid and histone‑fold proteins, and more broadly in other all‑α domains.	2.05	2
#10752	Short solvent-exposed connector loops	Generic structural signal for short, solvent-exposed loop/turn connectors between secondary structure elements—especially beta-beta hairpin loops and helix-strand junctions—enriched in small/charged residues (S/T, N, D/E, K/R); often adjacent to functional sites but not family-specific	2.03	2
#1888	AAA+/Lon noncatalytic residue peaks	Residue-level detector active in AAA+ ATPase / Lon-protease family proteins, with discrete peaks on structured non-catalytic positions and minimal activation on catalytic, ATP-binding, or disordered regions.	2.00	2
#2960	Helix N-cap pocket hotspots	Conserved strand-to-helix beginnings and helix N-cap segments that scaffold or border functional pockets—especially cofactor- and active-site regions—across many enzyme folds (PLP-, ThDP-, SAM-, and flavin-dependent) and, by extension, analogous short helical elements in non-enzymatic RNA/protein-binding proteins; these sites are enriched in small/hydrophobic residues with frequent glycine and acidic (Asp/Glu) capping positions.	1.90	2
#5818	Short amphipathic helix propensity	Short alpha-helical segment signal that is most prominent in acyl-CoA dehydrogenase family enzymes and a few other helical proteins, capturing local helix propensity and typical amphipathic/hydrophobic helical composition rather than specific catalytic or ligand-binding motifs.	1.79	2
#4608	BTB/POZ N-terminus beta-strand detector	Residue-level detector of beta-strand structural context in well-folded domains, with a strong preference for the very start of the BTB/POZ domain (the first beta-strand or its immediate N-terminal boundary). Outside BTB proteins it similarly marks isolated beta-strand (and adjacent inter-strand loop) positions within diverse domains rather than catalytic or ligand-coordination sites.	1.78	2
#2363	Charged alpha-helical tracts	Polar/charged alpha-helical tracts (often low-complexity): E/D/K/R/S/T/N/Q-rich, low-aromatic segments occurring both in intrinsically disordered regions and as solvent-exposed helices within folded domains; these amphipathic/coiled-coil–like helices mediate oligomerization, nucleic-acid/protein binding, and assembly, especially in viral proteins.	1.76	2
#1197	Exposed beta-loop binding rims	Short, solvent-exposed β-strand–to–loop segments that form ligand-recognition rims of β-rich domains—often glycan-binding in lectin/pentraxin folds and β-propeller CAZymes, but also protein-recognition loops in β-sandwich modules such as SPRY/PRY-SPRY, and analogous exposed loops near processing or substrate-binding sites in secreted virulence proteases—used in extracellular adhesion/innate-immunity proteins, secretory cargo lectins, and selected cytosolic adaptors; frequently adjacent to Ca2+-dependent sites where applicable.	1.74	2
#10995	β-strand–rich extracellular interfaces	β-strand–rich interaction surfaces with strong enrichment in secreted/lumenal protein regions, but not exclusive to them; the feature targets β-strands and adjacent loops that are often post-translationally stabilized (N-glycans) in extracellular proteins, and analogous β-elements in certain intracellular trafficking and phosphoinositide-processing proteins.	1.73	2
#15784	OmpR/PhoB signature motifs	Conserved sequence/structural elements of OmpR/PhoB-type response regulator transcription factors, marking signature positions in both the receiver domain and the C-terminal winged HTH DNA-binding domain.	1.73	2
#14603	Charged amphipathic interaction helices	Charged, amphipathic alpha‑helical interaction segments—often N‑terminal and coiled‑coil‑like—that mediate protein–protein or protein–membrane contacts in small complex subunits and diverse enzymes/membrane proteins across taxa	1.72	2
#2494	Structured-to-disordered boundary segments	A short, terminal or domain-edge coil/loop segment—most often a C‑terminal tail or a loop-to-helix/strand capping region—that marks the transition from a structured core to a flexible/low‑complexity extension. These segments frequently show reduced structural confidence (low pLDDT), are enriched in charged and glycine residues, and occasionally embed specific motifs (e.g., metal-binding cysteines), but typically lack discrete annotated active sites.	1.71	3
#13576	Amphipathic recognition helices	A broad detector of short, well-ordered amphipathic alpha-helices that serve as recognition/interaction surfaces (e.g., DNA-binding basic helices, kinase docking helices, coiled-coil edges), rather than catalytic motifs; consistent with a mixed basic/acidic and hydrophobic residue composition and active across diverse taxa and functions.	1.70	2
#7296	Bromodomains and flanking IDRs	Chromatin reader modules—especially bromodomains—and their immediately adjacent low-complexity, PTM-rich regulatory linkers/tails in nuclear chromatin regulators; the feature fires on the bromodomain core itself and extends into flanking acidic/Ser/Pro/Gly- and Lys/Arg-biased segments, with weaker signal on other reader domains (e.g., MBD and HMG-box) accompanied by similar IDRs; a minority of non-chromatin proteins with analogous acidic/basic low-complexity tracts can activate at lower scores (e.g., IAP E3 ligases, secreted phospholipases).	1.70	2
#7518	Histidine kinase helix-start recognition	Sparse recognition of residues within the cytosolic domains (PAS, HisKA/DHp, HATPase_c, REC) of two-component histidine kinases and related signal-transduction proteins, with peaks frequently landing on Pro or acidic (Asp/Glu) residues but also on a variety of other residues at helix-start and adjacent coil positions.	1.67	2
#6737	Cofactor adjacent functional segment	A contiguous, mid-protein "functional segment" used to position or interact with cofactors/ions or partner subunits—a loop/helix/strand adjacent to metal/ion or cofactor-binding sites (Fe–S, di-iron, Mg2+, Cu, SAM) or lying at assembly interfaces in redox/energy-metabolism complexes and related enzymes.	1.66	3
#8849	Alpha-helical CAZyme scaffolds	Alpha-helical scaffold segments in carbohydrate-active enzymes (glycoside hydrolases/lyases/phosphorylases): the feature highlights short runs within well-ordered core helices, often enriched in hydrophobic/aromatic residues, and typically avoids catalytic motifs and signal/transit peptides.	1.55	3
#11906	Post-signal N-terminal domain start	Order/disorder boundary and post-signal-peptide / domain-start segments in secreted and extracellular proteins: the feature activates on the immediately mature N-terminal portions of secreted/lumenal or extracellular proteins just after signal peptide cleavage, and more generally on flexible interdomain linkers or domain-start regions that transition from low-complexity/disordered sequence into structured elements (often initial beta strands), occasionally encompassing N-glycosylation sites.	1.52	2

Feature	Label	Description	Δ activation	Iso act.	Canon act.	Iso prev.	Canon prev.
#10704	N-terminal PEST-like IDRs	Low-complexity, often acidic/proline-rich intrinsically disordered N-terminal tails of intracellular eukaryotic proteins, frequently overlapping annotated PEST-like segments.	+5.24	10.09	4.84	18	8
#9998	Weak N+3 positional cue	Absolute N-terminal positional cue centered near the fourth residue (N+3), independent of amino-acid identity; most often within signal/transit peptides of precursors but also present in non-secreted proteins; the effect is weak and often does not produce a discrete detected site.	+4.07	11.70	7.62	3	2
#14234	Ordered cores and transmembrane helices	Generic preference for ordered chain regions of folded proteins, with particularly strong responses in hydrophobic α-helices that span membranes; avoids signal peptides and low-confidence N-terminal precursor regions.	+3.91	8.46	4.55	17	17
#12048	Generic secondary-structure context marker	A near-ubiquitous, low-amplitude feature marking generic local secondary-structure context—residues within α-helices, β-strands, and the connecting loops—with a frequent emphasis on helix–coil junctions and domain/linker regions, independent of residue type, function, or taxonomy.	+3.50	8.96	5.46	20	18
#8063	Disordered N-terminal activation window	Short N-terminal disordered segment found in a variety of proteins, including plant chloroplast transit peptides, bacterial prokaryotic ubiquitin-like protein Pup N-termini, anti-sigma factor N-tails, and some small RiPP precursor leaders. The activating region is Ser/Thr/Ala/Pro/Gly/Gln-rich, low in acidic residues, and typically lies in a disordered N-terminal stretch; the feature does not generally mark all transit peptides or all RiPP leaders.	+3.44	7.48	4.05	11	8
#13334	N-terminal IDR boundary	Recognition of short, low-complexity, intrinsically disordered N-terminal tails immediately upstream of the first folded domain; these segments are often Ser/Thr/Pro/Gly/Ala-biased and sometimes extend a few residues into the initial helix or the very beginning of the first domain (domain boundary/disorder-to-order transition), especially in transcription regulators and co-regulators.	+3.33	7.79	4.46	42	17
#6854	Low complexity helical repeats	A general, composition-driven signal for non-globular sequence regions: it prefers low-complexity, compositionally biased tracts and repetitive helical polymers, emphasizing charged/polar (E, D, K, R, S, Q) or Gly/Pro-rich content, and residues at helix–coil transitions; it is largely muted in compact, well-packed helical cores.	+3.26	5.69	2.43	28	12
#10977	Early N-terminal disordered patch	Short N-terminal patch at the extreme N-terminus (around residues 5–10), capturing the first distinctive segment of preproteins/precursors and N-terminal disordered tails; residue composition is heterogeneous but frequently polar (Ser/Thr) and/or charged.	+3.13	7.84	4.71	7	3
#13449	Membrane-biased alpha-helix detector	Generic alpha-helix detector with strongest preference for long hydrophobic helices that associate with membranes (transmembrane segments and signal peptides), but also activating on extended coiled-coils and internal alpha-helices in soluble proteins	+2.77	8.12	5.35	15	12
#8015	Short helical docking motifs	Short protein–protein interaction segments in eukaryotic intracellular signaling scaffolds and regulators; includes amphipathic docking/dimerization helices as well as interaction patches embedded within small folded signaling domains (e.g., RA, DH, PH), exemplified by AKAP PKA regulatory subunit–binding helices and extending to BH3, bHLH/coiled-coil, and other helical docking elements across diverse signaling, cytoskeletal, and transcriptional complexes.	+2.69	6.73	4.05	12	8
#5437	Coiled-coils and propeptide spacers	Long α-helical coiled-coil regions of large scaffold/structural proteins, plus the propeptide/spacer regions of secreted neuropeptide and prohormone precursors. The feature responds to heptad-repeat coiled-coil sequence character (typically with R/K-X-X-L/E patterns and charged residues in a/d/e/g positions) and to short polar/charged propeptide segments flanking dibasic cleavage sites.	+2.52	4.11	1.59	13	2
#2895	Motif-and-repeat sequence detector	Sequence-pattern detector that responds most strongly to bacterial helix-turn-helix DNA-binding motifs, particularly the C-terminal HTH of sigma-54-dependent response regulators and related HTH-type transcription factors, and also to (i) dibasic Lys/Arg convertase cleavage motifs and residues flanking amidated peptides in secreted prohormone/peptide precursors, (ii) Gly–X–Y/proline-rich collagen-like repeats (including hydroxyproline sites), (iii) hydrophobic a/d positions in coiled-coil heptads of long scaffolds, and (iv) PTM-prone Ser/Thr embedded in Lys/Arg-rich disordered tails of chromatin/DNA-binding proteins. The unifying concept is recognition of short DNA-binding/processing/PTM motifs and repetitive/low-complexity sequence context used across diverse proteins.	+2.23	4.57	2.34	12	7
#15815	Basic N-terminal targeting patch	Short, basic/polar N‑terminal leader/transit segment immediately after the initiator methionine (positions ~3–6) — i.e., the positively charged/polar start of signal peptides and organelle transit peptides, and more generally low‑acidic, disordered N‑terminal patches (including occasional cysteine‑rich starts) used for targeting, export, rRNA binding, or membrane topology cues	+2.23	10.90	8.67	7	4
#11526	Glycine amidation motif detector	Residue-level detector for small/flexible residues—especially glycine—in short, low-structure linkers and proteolytic processing signals of peptide precursors, with a strong preference for the C‑terminal amidation context in which a glycine (amide donor) immediately precedes mono/di‑basic residues (G‑K/R); outside precursors it gives sparse hits on similar small-residue sites in intrinsically disordered regions and occasionally within structured domains across diverse taxa.	+2.17	4.33	2.17	13	7
#4834	DNA-binding and helical repeats	General helical secondary-structure elements, with strongest activation on short DNA-binding recognition helices (HTH/homeodomain/σ-factor region 4) and on extended helical repeats including collagen Gly–X–Y triple helices; also frequently engages signal/transmembrane helices in secreted and membrane proteins	+2.14	4.73	2.58	12	4
#612	Proline-rich disordered regulatory linkers	Intrinsically disordered regulatory linkers and targeting segments that flank signaling/catalytic domains—most prominently adjacent to RhoGAP domains—enriched in proline/serine/threonine/acidic/glycine residues and often containing SH3-binding PxxP motifs and/or polybasic membrane/focal-adhesion–targeting patches; in some cases the activation extends into the N-terminal edge of the RhoGAP catalytic domain (including the arginine-finger vicinity). Similar sequence/structural signatures are also recognized in flexible interdomain hinges of other eukaryotic signaling proteins.	+2.07	4.57	2.49	41	8
#7565	Hydrophobic helices and signals	Generic, low-specificity signal for short hydrophobic/alpha-helical stretches with a mild N-terminal bias, encompassing signal/transit peptides, single-pass transmembrane anchors, and coiled-coil helices; also yields weak single-residue responses to hydrophobic/aromatic positions within structured loops and peptide-hormone segments.	+2.06	7.83	5.77	16	13
#1722	S/T/Pro-rich disordered regions	Intrinsically disordered, low-complexity sequence elements enriched in Ser/Thr/Pro/polar residues, characteristic of flexible linkers, regulatory tails, and polar/low-complexity tracts across diverse taxa.	+2.04	5.66	3.62	12	5
#5321	N-terminal domain-start leader regions	N-terminal leader/capping segments at the start of a protein or of a new domain, including pre-/pro-peptides, signal-anchor transmembrane helices with their flanking helices, and the first β/α elements that initiate the structured core; the feature emphasizes positional “domain/start-of-mature-chain” regions rather than catalytic metal-binding cores.	+1.98	3.68	1.70	34	4
#3604	Acidic Pro/Ser-rich disordered regions	Proline/serine-rich low-complexity and disordered regions, frequently with acidic/PEST-like character; the feature also fires on small proteins and on internal stretches outside obvious low-complexity tracts when local Pro/Arg/Ser content is high.	+1.91	4.09	2.18	6	4
#11395	Spire KIND/WH2 with charged LCRs	Conserved motifs within the KIND domain and flanking the WH2 actin-binding motifs of Spire family actin nucleators, together with low-complexity, intrinsically disordered regions enriched in charged/polar residues (E/D/S/T/K/R) in other proteins; occasionally shows weaker signal within or flanking helical segments (e.g., transmembrane or coiled-coil boundaries).	+1.86	5.64	3.78	9	7
#9044	General helical structural elements	General helical structural elements: the feature marks residues within helical segments across many protein classes—alpha-helices in HTH/homeodomain DNA-binding regions, coiled-coils, transmembrane helices, and signal-peptide/propeptide helices—with a characteristic periodic emphasis on Gly in the collagen triple-helix (Gly–X–Y) repeat.	+1.76	3.86	2.09	11	7
#2257	Eukaryotic low-complexity IDRs	Intrinsic disorder/low-complexity signal: the feature marks long, compositionally biased IDRs in eukaryotic proteins—especially Lys/Arg-rich basic patches, Glu/Asp-rich acidic tracts, and Q/N/S/T-rich stretches—found in N-/C-terminal tails and inter-domain linkers, while avoiding well-folded domains and transmembrane helices; commonly seen in translation/RNA biogenesis factors and secretory/quality-control proteins.	+1.76	3.35	1.59	27	4
#6452	Chlorophyll/heme NGRAAM/WxWG motifs	A sequence-motif feature firing on short conserved patches in chlorophyll/heme-related proteins, including light-harvesting-like (LHC/HLI/OHP) proteins and ferrochelatases, with recurrent context around an "NGRAAM"-like motif and an aromatic "WxWG" patch; also activates on short patches in small membrane-regulator peptides.	+1.60	4.86	3.26	19	17
#1370	Short low-complexity N-terminal segment	Short, low‑complexity, intrinsically disordered N‑terminal segments (~10–20 aa) that frequently correspond to leader/transit/propeptide presequences (e.g., chloroplast transit peptides, RiPP/bacteriocin leaders) or generic unstructured N‑tails; typically non‑hydrophobic and enriched in Ser/Thr/Pro but tolerant of basic (poly‑Arg/Lys) or Cys‑rich compositions	+1.58	4.16	2.58	12	10
#12004	N-terminal Pro/His helical motifs	Short N-terminal sequence motifs in metabolic enzymes, characterized by Pro- and His-containing patches embedded in early α-helical segments (e.g. S-P-x-P-Y/F-H in aspartyl aminopeptidases; G-H-P-D in S-adenosylmethionine synthases)	+1.57	3.73	2.16	6	2
#14712	SPS N-terminal regulatory motif	A conserved N-terminal regulatory motif in plant sucrose-phosphate synthases (SPS), centered on a short ATR(N/S/P)xRERS / SPQERN-type sequence upstream of the catalytic glycosyltransferase core.	+1.55	5.53	3.98	16	11
#1803	Unknown generic feature	Unknown generic feature	+6.04	17.75	11.71	174	143
#14534	Unknown generic feature	Unknown generic feature	+3.75	15.23	11.48	173	143
#9005	Unknown generic feature	Unknown generic feature	+2.38	12.15	9.77	142	124

Feature	Label	Description	Activation	Prevalence
#5725	UBC/E2-like fold recognition	UBC/E2-like fold recognition across ubiquitin and ubiquitin-like conjugation systems, capturing catalytically active E2s and inactive homologs (UEV, RWD) and the cognate ESCRT components that harbor them; the signal corresponds to the conserved UBC-core surface, with a bias toward the N-terminal α-helix, and can extend to physicochemically similar acidic, α-helical blocks that mimic this surface	17.52	31
#4992	Small-residue disordered N-terminus	Context-dependent free N-terminus signature: short, disordered amino-terminal stretches beginning with the initiator Met followed by residues compatible with MAP cleavage or N-acetylation (e.g., M–G/A/S/T/V, and other contexts including M–D/E/L); strongest emphasis on small residues at positions 2–3; activation is absent when the native N-terminus is structured or starts with a signal peptide/transmembrane segment.	15.46	4
#15815	Basic N-terminal targeting patch	Short, basic/polar N‑terminal leader/transit segment immediately after the initiator methionine (positions ~3–6) — i.e., the positively charged/polar start of signal peptides and organelle transit peptides, and more generally low‑acidic, disordered N‑terminal patches (including occasional cysteine‑rich starts) used for targeting, export, rRNA binding, or membrane topology cues	10.90	4
#10704	N-terminal PEST-like IDRs	Low-complexity, often acidic/proline-rich intrinsically disordered N-terminal tails of intracellular eukaryotic proteins, frequently overlapping annotated PEST-like segments.	10.09	13
#5241	Early N-terminal positional signal	Generic early N-terminus positional signal peaking at residue ~5–7, found broadly within the disordered N-tail of proteins, including both cleavable leader/targeting segments (signal peptides, propeptides) of precursor proteins and disordered N-terminal regions of mature proteins; residue identity is not specific	9.59	2
#1066	UBC/UEV HPN-loop interface	Strand-helix-loop interaction patch of compact alpha/beta domains, most prominently the UBC/UEV (ubiquitin-conjugating-like) fold where it captures the conserved HPN-loop and adjacent beta-strand/helix that form the partner-binding surface; secondarily, the feature recognizes geometrically analogous beta-strand–helix patches in other interaction/DNA-binding domains (e.g., T-box and EVH1-like).	9.14	30
#12048	Generic secondary-structure context marker	A near-ubiquitous, low-amplitude feature marking generic local secondary-structure context—residues within α-helices, β-strands, and the connecting loops—with a frequent emphasis on helix–coil junctions and domain/linker regions, independent of residue type, function, or taxonomy.	8.96	4
#14234	Ordered cores and transmembrane helices	Generic preference for ordered chain regions of folded proteins, with particularly strong responses in hydrophobic α-helices that span membranes; avoids signal peptides and low-confidence N-terminal precursor regions.	8.46	3
#13449	Membrane-biased alpha-helix detector	Generic alpha-helix detector with strongest preference for long hydrophobic helices that associate with membranes (transmembrane segments and signal peptides), but also activating on extended coiled-coils and internal alpha-helices in soluble proteins	8.12	4
#10977	Early N-terminal disordered patch	Short N-terminal patch at the extreme N-terminus (around residues 5–10), capturing the first distinctive segment of preproteins/precursors and N-terminal disordered tails; residue composition is heterogeneous but frequently polar (Ser/Thr) and/or charged.	7.84	4
#7565	Hydrophobic helices and signals	Generic, low-specificity signal for short hydrophobic/alpha-helical stretches with a mild N-terminal bias, encompassing signal/transit peptides, single-pass transmembrane anchors, and coiled-coil helices; also yields weak single-residue responses to hydrophobic/aromatic positions within structured loops and peptide-hormone segments.	7.83	4
#13334	N-terminal IDR boundary	Recognition of short, low-complexity, intrinsically disordered N-terminal tails immediately upstream of the first folded domain; these segments are often Ser/Thr/Pro/Gly/Ala-biased and sometimes extend a few residues into the initial helix or the very beginning of the first domain (domain boundary/disorder-to-order transition), especially in transcription regulators and co-regulators.	7.79	31
#8063	Disordered N-terminal activation window	Short N-terminal disordered segment found in a variety of proteins, including plant chloroplast transit peptides, bacterial prokaryotic ubiquitin-like protein Pup N-termini, anti-sigma factor N-tails, and some small RiPP precursor leaders. The activating region is Ser/Thr/Ala/Pro/Gly/Gln-rich, low in acidic residues, and typically lies in a disordered N-terminal stretch; the feature does not generally mark all transit peptides or all RiPP leaders.	7.48	9
#13818	N-terminal domain start signal	Start-of-domain signal: the feature detects the transition into the first ordered secondary-structure elements at the N-terminus of a conserved globular domain (often immediately after a signal/transit peptide, lipidation site, or disordered tail), with strongest matches for cyclophilin-type PPIase folds and weaker matches for many other domain starts.	7.44	31
#8015	Short helical docking motifs	Short protein–protein interaction segments in eukaryotic intracellular signaling scaffolds and regulators; includes amphipathic docking/dimerization helices as well as interaction patches embedded within small folded signaling domains (e.g., RA, DH, PH), exemplified by AKAP PKA regulatory subunit–binding helices and extending to BH3, bHLH/coiled-coil, and other helical docking elements across diverse signaling, cytoskeletal, and transcriptional complexes.	6.73	4
#7986	Intrinsically disordered low-complexity regions	Compositionally biased, intrinsically disordered low‑complexity regions used as flexible interaction/nucleic‑acid–binding modules—enriched in Pro/Ser/Lys/Glu/Gln/Thr/Gly/Arg and encompassing poly‑Pro tracts, Lys/Arg‑rich basic stretches, and acidic Asp/Glu clusters—common in transcription/RNA‑processing factors, viral basic regulators, and the C‑terminal scaffolds of RNase E	6.68	22
#16243	Structured-region leucine detector	Generic detector of leucine side chains in structured regions of folded domains, with occasional weaker responses to other hydrophobics; largely indifferent to specific functional motifs or protein class.	6.68	2
#8237	Conserved structured domain peaks	A feature that activates within structured partner-recognition/DNA-binding domains and adjacent ordered segments of eukaryotic regulatory proteins, with peaks falling on conserved residues inside these domains rather than on flanking low-complexity stretches.	6.67	2
#14927	Structural and domain boundary residues	Residues at structural and domain junctions: starts/ends of secondary-structure elements (especially helix starts and transmembrane-helix termini) and positions immediately preceding or following annotated domains, often adjacent to ligand/cofactor-binding regions	6.45	2
#7026	Catalytic domain helix/loop boundary motif	A localized internal-domain motif that marks a specific short helix/loop boundary within a catalytic domain, situated downstream of the catalytic proton-donor residue in α/β hydrolase-like folds.	6.38	2
#934	SprT metalloprotease dual-motif detector	Detector of conserved sequence motifs within SprT-like metalloprotease domains, particularly in SPRTN orthologs and Germ cell nuclear acidic protein (GCNA) family proteins.	6.32	2
#9237	Glycine detector in disordered regions	Residue-identity detector for glycine (G), with a mild bias toward intrinsically disordered/low-complexity segments (often including very N-terminal residues); pan-taxonomic and function-agnostic, but activation is sparse and frequently absent, becoming detectable mainly in glycine-rich low-complexity contexts; occasional weak responses to other small/polar residues.	6.14	3
#13480	Alanine and small-residue enrichment	A sequence-composition feature that detects small, non-aromatic residues—especially alanine—and their enrichment, lighting up individual Ala residues and stretches enriched in A/G/S/T/P/V (alanine-rich, low-complexity segments) irrespective of protein function or secondary structure.	6.07	4
#627	Generic serine detector	Generic serine detector: the feature marks serine residues, with weaker affinity for threonine (and occasionally proline), and is especially prominent in low-complexity S/T-rich stretches; it is context- and structure-agnostic and does not correspond to a specific functional motif.	6.05	4
#14895	Unknown generic feature	Unknown generic feature	21.36	31
#9214	Unknown generic feature	Unknown generic feature	18.11	31
#1803	Unknown generic feature	Unknown generic feature	17.75	31
#9194	Unknown generic feature	Unknown generic feature	15.70	31
#14534	Unknown generic feature	Unknown generic feature	15.23	31
#9005	Unknown generic feature	Unknown generic feature	12.15	31

Clinical variants

Differential region — N-terminal extension (isoform-unique)

Pos (iso)	AA change	Consequence	Source	Clin. sig.	AF (gnomAD)	Impact	AlphaMissense	ESM-C ΔLLR	Link
0	I→I	synonymous_variant	gnomAD	—	3.35e-06	—	N/A	—	chr5-139561701-C-T
2	E→A	missense_variant	gnomAD	—	1.07e-06	—	N/A	2.60	chr5-139561706-A-C
2	E→D	missense_variant	gnomAD	—	1.05e-06	—	N/A	-0.89	chr5-139561707-G-C
2	E→D	missense_variant	gnomAD	—	1.05e-06	—	N/A	-0.89	chr5-139561707-G-T
3	A→G	missense_variant	gnomAD	—	1.04e-06	—	N/A	-0.27	chr5-139561709-C-G
4	A→T	missense_variant	gnomAD	—	6.06e-06	—	N/A	-1.62	chr5-139561711-G-A
4	A→D	missense_variant	gnomAD	—	9.89e-07	—	N/A	-2.11	chr5-139561712-C-A
4	A→A	synonymous_variant	gnomAD	—	1.92e-06	—	N/A	0.00	chr5-139561713-C-G
4	A→A	synonymous_variant	gnomAD	—	1.63e-05	—	N/A	0.00	chr5-139561713-C-T
5	S→P	missense_variant	gnomAD	—	1.91e-06	—	N/A	0.70	chr5-139561714-T-C
5	S→S	synonymous_variant	gnomAD	—	9.50e-07	—	N/A	0.00	chr5-139561716-A-C
5	S→S	synonymous_variant	gnomAD	—	8.55e-06	—	N/A	0.00	chr5-139561716-A-G
6	G→S	missense_variant	gnomAD	—	9.29e-07	—	N/A	0.45	chr5-139561717-G-A
6	G→D	missense_variant	gnomAD	—	5.02e-05	—	N/A	-2.19	chr5-139561718-G-A
6	G→V	missense_variant	gnomAD	—	9.30e-07	—	N/A	-0.69	chr5-139561718-G-T
7	—	frameshift_variant	gnomAD	—	9.16e-07	LoF	—	—	chr5-139561720-T-TC
7	S→F	missense_variant	gnomAD	—	7.59e-05	—	N/A	-1.64	chr5-139561721-C-T
7	S→S	synonymous_variant	gnomAD	—	8.80e-07	—	N/A	0.00	chr5-139561722-C-T
8	L→L	synonymous_variant	gnomAD	—	3.27e-05	—	N/A	0.00	chr5-139561725-A-G
9	A→T	missense_variant	gnomAD	—	1.74e-06	—	N/A	-1.50	chr5-139561726-G-A
9	A→S	missense_variant	gnomAD	—	6.10e-06	—	N/A	0.91	chr5-139561726-G-T
9	—	frameshift_variant	gnomAD	—	8.65e-07	LoF	—	—	chr5-139561727-C-CA
9	—	frameshift_variant	gnomAD	—	2.42e-05	LoF	—	—	chr5-139561727-C-CCCCTTCCCCGT
9	A→V	missense_variant	gnomAD	—	3.46e-06	—	N/A	-1.64	chr5-139561727-C-T
9	—	frameshift_variant	gnomAD	—	2.68e-05	LoF	—	—	chr5-139561727-CCCCTTCCCCGT-C
9	A→A	synonymous_variant	gnomAD	—	8.58e-07	—	N/A	0.00	chr5-139561728-C-A
9	A→A	synonymous_variant	gnomAD	—	1.72e-06	—	N/A	0.00	chr5-139561728-C-G
10	P→T	missense_variant	gnomAD	—	2.55e-06	—	N/A	-2.18	chr5-139561729-C-A
10	P→S	missense_variant	gnomAD	—	1.70e-06	—	N/A	-0.71	chr5-139561729-C-T
10	P→P	synonymous_variant	gnomAD	—	1.69e-06	—	N/A	0.00	chr5-139561731-T-C
11	—	frameshift_variant	gnomAD	—	2.51e-06	LoF	—	—	chr5-139561732-TCCCCGTCCCTTCCCCGC-T
11	S→S	synonymous_variant	gnomAD	—	3.30e-06	—	N/A	0.00	chr5-139561734-C-T
12	P→A	missense_variant	gnomAD	—	2.46e-06	—	N/A	-0.88	chr5-139561735-C-G
12	P→S	missense_variant	gnomAD	—	8.21e-07	—	N/A	-0.86	chr5-139561735-C-T
12	P→R	missense_variant	gnomAD	—	3.29e-06	—	N/A	-0.72	chr5-139561736-C-G
12	P→L	missense_variant	gnomAD	—	8.05e-05	—	N/A	-0.50	chr5-139561736-C-T
12	P→P	synonymous_variant	gnomAD	—	1.82e-06	—	N/A	0.00	chr5-139561737-G-A
12	P→P	synonymous_variant	gnomAD	—	9.11e-07	—	N/A	0.00	chr5-139561737-G-T
13	S→T	missense_variant	gnomAD	—	8.67e-07	—	N/A	-2.00	chr5-139561738-T-A
13	S→A	missense_variant	gnomAD	—	9.11e-05	—	N/A	0.20	chr5-139561738-T-G
13	—	frameshift_variant	gnomAD	—	3.47e-06	LoF	—	—	chr5-139561738-T-TC
13	—	frameshift_variant	gnomAD	—	8.67e-07	LoF	—	—	chr5-139561738-T-TG
13	—	inframe_deletion	gnomAD	—	7.81e-06	—	N/A	—	chr5-139561738-TCCCTTCCCCGCC-T
13	S→F	missense_variant	gnomAD	—	8.39e-07	—	N/A	-1.56	chr5-139561739-C-T
13	S→S	synonymous_variant	COSMIC	—	—	—	N/A	0.00	COSV54077651
14	—	frameshift_variant	gnomAD	—	3.11e-05	LoF	—	—	chr5-139561741-C-CA
14	—	frameshift_variant	gnomAD	—	3.75e-05	LoF	—	—	chr5-139561741-C-CG
14	—	frameshift_variant	gnomAD	—	7.98e-07	LoF	—	—	chr5-139561741-C-CT
14	L→F	missense_variant	gnomAD	—	7.98e-07	—	N/A	-1.80	chr5-139561741-C-T
14	L→P	missense_variant	gnomAD	—	2.80e-06	—	N/A	1.44	chr5-139561742-T-C
14	L→R	missense_variant	gnomAD	—	1.11e-04	—	N/A	1.66	chr5-139561742-T-G
14	—	frameshift_variant	gnomAD	—	1.77e-05	LoF	—	—	chr5-139561742-TTCCCCGCCCCC-T
14	—	inframe_deletion	gnomAD	—	1.01e-04	—	N/A	—	chr5-139561742-TTCCCCGCCCCCG-T
14	L→L	synonymous_variant	gnomAD	—	1.00e-05	—	N/A	0.00	chr5-139561743-T-A
14	L→L	synonymous_variant	gnomAD	—	2.25e-05	—	N/A	0.00	chr5-139561743-T-C
14	—	frameshift_variant	gnomAD	—	2.39e-04	LoF	—	—	chr5-139561743-T-TA
14	—	frameshift_variant	gnomAD	—	2.35e-04	LoF	—	—	chr5-139561743-T-TC
14	—	frameshift_variant	gnomAD	—	1.79e-04	LoF	—	—	chr5-139561743-T-TG
14	—	frameshift_variant	gnomAD	—	8.34e-07	LoF	—	—	chr5-139561743-T-TTCCC
14	—	frameshift_variant	gnomAD	—	5.84e-06	LoF	—	—	chr5-139561743-TCCCC-T
14	—	inframe_deletion	gnomAD	—	6.67e-06	—	N/A	—	chr5-139561743-TCCCCGC-T
14	—	frameshift_variant	gnomAD	—	8.34e-07	LoF	—	—	chr5-139561743-TCCCCGCCCCCGTCCCCG-T
15	P→S	missense_variant	gnomAD	—	9.21e-06	—	N/A	-0.91	chr5-139561744-C-T
15	P→H	missense_variant	gnomAD	—	1.59e-06	—	N/A	-2.97	chr5-139561745-C-A
15	P→R	missense_variant	gnomAD	—	1.59e-06	—	N/A	-0.58	chr5-139561745-C-G
15	P→L	missense_variant	gnomAD	—	2.38e-04	—	N/A	-1.36	chr5-139561745-C-T
15	P→P	synonymous_variant	gnomAD	—	1.57e-06	—	N/A	0.00	chr5-139561746-C-A
15	P→P	synonymous_variant	gnomAD	—	7.87e-07	—	N/A	0.00	chr5-139561746-C-G
15	P→P	synonymous_variant	gnomAD	—	3.54e-05	—	N/A	0.00	chr5-139561746-C-T
16	R→S	missense_variant	gnomAD	—	1.58e-06	—	N/A	-0.88	chr5-139561747-C-A
16	R→C	missense_variant	gnomAD	—	1.58e-06	—	N/A	-2.38	chr5-139561747-C-T
16	—	frameshift_variant	gnomAD	—	8.06e-05	LoF	—	—	chr5-139561747-CG-C
16	R→H	missense_variant	gnomAD	—	5.08e-06	—	N/A	-2.27	chr5-139561748-G-A
16	—	frameshift_variant	gnomAD	—	2.03e-05	LoF	—	—	chr5-139561748-G-GC
16	R→L	missense_variant	gnomAD	—	2.03e-05	—	N/A	-0.65	chr5-139561748-G-T
16	—	frameshift_variant	gnomAD	—	2.54e-06	LoF	—	—	chr5-139561748-GC-G
16	—	frameshift_variant	gnomAD	—	2.54e-06	LoF	—	—	chr5-139561748-GCC-G
16	R→R	synonymous_variant	gnomAD	—	7.87e-07	—	N/A	0.00	chr5-139561749-C-A
16	R→R	synonymous_variant	gnomAD	—	1.42e-05	—	N/A	0.00	chr5-139561749-C-T
16	—	frameshift_variant	COSMIC	—	—	LoF	—	—	COSV99551267
17	P→A	missense_variant	gnomAD	—	7.79e-07	—	N/A	-1.25	chr5-139561750-C-G
17	P→S	missense_variant	gnomAD	—	3.97e-05	—	N/A	-1.87	chr5-139561750-C-T
17	P→R	missense_variant	gnomAD	—	1.55e-06	—	N/A	-0.91	chr5-139561751-C-G
17	P→P	synonymous_variant	gnomAD	—	1.54e-06	—	N/A	0.00	chr5-139561752-C-G
18	R→S	missense_variant	gnomAD	—	7.68e-07	—	N/A	-1.22	chr5-139561753-C-A
18	R→G	missense_variant	gnomAD	—	7.68e-07	—	N/A	-0.88	chr5-139561753-C-G
18	R→C	missense_variant	gnomAD	—	1.54e-06	—	N/A	-2.36	chr5-139561753-C-T
18	—	frameshift_variant	gnomAD	—	1.54e-06	LoF	—	—	chr5-139561753-CG-C
18	—	inframe_deletion	gnomAD	—	7.68e-07	—	N/A	—	chr5-139561753-CGTCCCCGCCCCG-C
18	R→H	missense_variant	gnomAD	—	2.33e-06	—	N/A	-2.33	chr5-139561754-G-A
18	R→L	missense_variant	gnomAD	—	1.40e-05	—	N/A	-1.20	chr5-139561754-G-T
18	—	frameshift_variant	gnomAD	—	6.98e-06	LoF	—	—	chr5-139561754-GT-G
18	R→R	synonymous_variant	gnomAD	—	3.18e-06	—	N/A	0.00	chr5-139561755-T-A
18	R→R	synonymous_variant	gnomAD	—	2.13e-04	—	N/A	0.00	chr5-139561755-T-C
18	R→R	synonymous_variant	gnomAD	—	4.52e-04	—	N/A	0.00	chr5-139561755-T-G
18	—	frameshift_variant	gnomAD	—	2.33e-05	LoF	—	—	chr5-139561755-T-TC
18	—	frameshift_variant	gnomAD	—	2.12e-06	LoF	—	—	chr5-139561755-TC-T
19	P→S	missense_variant	gnomAD	—	1.56e-06	—	N/A	-1.66	chr5-139561756-C-T
19	P→P	synonymous_variant	gnomAD	—	7.48e-07	—	N/A	0.00	chr5-139561758-C-T
20	R→G	missense_variant	gnomAD	—	1.50e-06	—	N/A	-0.47	chr5-139561759-C-G
20	R→C	missense_variant	gnomAD	—	7.49e-07	—	N/A	-2.55	chr5-139561759-C-T
20	R→H	missense_variant	gnomAD	—	4.60e-05	—	N/A	-2.43	chr5-139561760-G-A
20	R→L	missense_variant	gnomAD	—	7.66e-06	—	N/A	-1.40	chr5-139561760-G-T
20	—	frameshift_variant	gnomAD	—	2.55e-06	LoF	—	—	chr5-139561760-GC-G
20	R→R	synonymous_variant	gnomAD	—	7.48e-07	—	N/A	0.00	chr5-139561761-C-A
20	R→R	synonymous_variant	gnomAD	—	1.50e-06	—	N/A	0.00	chr5-139561761-C-G
20	R→R	synonymous_variant	gnomAD	—	1.50e-06	—	N/A	0.00	chr5-139561761-C-T
21	P→S	missense_variant	gnomAD	—	1.49e-06	—	N/A	-1.27	chr5-139561762-C-T
21	—	frameshift_variant	gnomAD	—	2.98e-06	LoF	—	—	chr5-139561764-C-CG
21	P→P	synonymous_variant	gnomAD	—	2.24e-06	—	N/A	0.00	chr5-139561764-C-G
21	P→P	synonymous_variant	gnomAD	—	7.45e-07	—	N/A	0.00	chr5-139561764-C-T
21	—	frameshift_variant	gnomAD	—	3.73e-06	LoF	—	—	chr5-139561764-CG-C
22	G→E	missense_variant	gnomAD	—	1.20e-04	—	N/A	-1.92	chr5-139561766-G-A
23	G→D	missense_variant	gnomAD	—	2.24e-06	—	N/A	-1.42	chr5-139561769-G-A
23	G→G	synonymous_variant	gnomAD	—	3.02e-06	—	N/A	0.00	chr5-139561770-C-A
24	R→H	missense_variant	gnomAD	—	1.04e-06	—	N/A	-1.71	chr5-139561772-G-A
24	R→R	synonymous_variant	gnomAD	—	7.44e-07	—	N/A	0.00	chr5-139561773-C-T
25	R→C	missense_variant	gnomAD	—	7.44e-07	—	N/A	-2.36	chr5-139561774-C-T
25	R→H	missense_variant	gnomAD	—	1.01e-06	—	N/A	-2.17	chr5-139561775-G-A
25	—	frameshift_variant	gnomAD	—	2.03e-06	LoF	—	—	chr5-139561775-G-GCCACCCGCCTC
25	R→L	missense_variant	gnomAD	—	1.01e-06	—	N/A	-1.19	chr5-139561775-G-T
26	—	frameshift_variant	gnomAD	—	9.97e-07	LoF	—	—	chr5-139561778-AC-A
26	H→Q	missense_variant	gnomAD	—	7.42e-06	—	N/A	0.39	chr5-139561779-C-G
26	H→H	synonymous_variant	gnomAD	—	9.65e-06	—	N/A	0.00	chr5-139561779-C-T
27	P→A	missense_variant	gnomAD	—	7.42e-07	—	N/A	-1.67	chr5-139561780-C-G
27	P→S	missense_variant	gnomAD	—	1.48e-06	—	N/A	-1.33	chr5-139561780-C-T
27	P→R	missense_variant	gnomAD	—	1.48e-06	—	N/A	-0.09	chr5-139561781-C-G
27	P→P	synonymous_variant	gnomAD	—	2.18e-06	—	N/A	0.00	chr5-139561782-G-A
27	P→P	synonymous_variant	gnomAD	—	1.09e-06	—	N/A	0.00	chr5-139561782-G-T
28	P→S	missense_variant	gnomAD	—	1.48e-06	—	N/A	-1.09	chr5-139561783-C-T
28	P→R	missense_variant	gnomAD	—	4.44e-05	—	N/A	-0.31	chr5-139561784-C-G
28	P→L	missense_variant	gnomAD	—	1.48e-06	—	N/A	-1.26	chr5-139561784-C-T
29	P→S	missense_variant	gnomAD	—	1.48e-06	—	N/A	-1.24	chr5-139561786-C-T
29	P→R	missense_variant	gnomAD	—	7.40e-07	—	N/A	-0.68	chr5-139561787-C-G
29	P→L	missense_variant	gnomAD	—	7.40e-07	—	N/A	-1.54	chr5-139561787-C-T
29	P→P	synonymous_variant	gnomAD	—	1.33e-05	—	N/A	0.00	chr5-139561788-C-T
29	P→S	missense_variant	COSMIC	—	—	—	N/A	-1.24	COSV99551256
30	T→A	missense_variant	gnomAD	—	8.87e-07	—	N/A	1.28	chr5-139561789-A-G
30	T→T	synonymous_variant	gnomAD	—	2.95e-06	—	N/A	0.00	chr5-139561791-C-T

139 variants in the differential region.

Shared canonical core — sequence common to canonical and isoform; AlphaMissense applies here

Pos (iso)	AA change	Consequence	Source	Clin. sig.	AF (gnomAD)	Impact	AlphaMissense	ESM-C ΔLLR	Link
31	M→L	missense_variant	gnomAD	—	3.29e-06	damaging	—	-9.87	chr5-139561792-A-C
31	M→V	missense_variant	gnomAD	—	1.64e-06	damaging	—	-10.50	chr5-139561792-A-G
31	—	frameshift_variant	gnomAD	—	8.96e-07	LoF	—	—	chr5-139561793-TG-T
32	A→S	missense_variant	gnomAD	—	7.37e-07	damaging	ambiguous (0.45)	-7.84	chr5-139561795-G-T
32	A→A	synonymous_variant	gnomAD	—	8.36e-06	—	—	0.00	chr5-139561797-T-G
33	L→L	synonymous_variant	gnomAD	—	3.61e-05	—	—	0.00	chr5-139561798-C-T
34	K→E	missense_variant	gnomAD	—	7.41e-07	damaging	likely_pathogenic (0.94)	-9.31	chr5-139561801-A-G
34	K→K	synonymous_variant	gnomAD	—	7.44e-07	—	—	0.00	chr5-139561803-G-A
35	R→R	synonymous_variant	gnomAD	—	7.40e-07	—	—	0.00	chr5-139561804-A-C
36	I→I	synonymous_variant	gnomAD	—	7.54e-07	—	—	0.00	chr5-139561809-C-T
37	H→H	synonymous_variant	gnomAD	—	7.39e-07	—	—	0.00	chr5-139561812-C-T
37	H→P	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.88)	-11.36	COSV105026649
38	K→E	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.95)	-10.50	COSV54077432
38	K→K	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV105026650
39	E→Q	missense_variant	ClinVar	—	—	damaging	likely_pathogenic (0.99)	-11.25	ClinVar:4311627
39	E→Q	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-11.25	COSV99551389
41	N→N	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600380-T-C
42	D→D	synonymous_variant	gnomAD	—	1.37e-06	—	—	0.00	chr5-139600383-T-C
43	L→L	synonymous_variant	gnomAD	—	3.28e-05	—	—	0.00	chr5-139600384-C-T
43	L→L	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600386-G-C
44	A→A	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600389-A-G
45	R→G	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.67)	-10.06	chr5-139600390-C-G
45	R→W	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.82)	-11.43	chr5-139600390-C-T
45	R→R	synonymous_variant	gnomAD	—	2.74e-06	—	—	0.00	chr5-139600392-G-A
45	R→Q	missense_variant	COSMIC	—	—	damaging	ambiguous (0.37)	-8.31	COSV54079740
46	D→D	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600395-C-T
48	P→P	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV54077679
49	A→S	missense_variant	gnomAD	—	6.84e-07	—	likely_benign (0.15)	-5.78	chr5-139600402-G-T
49	A→A	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600404-A-G
49	A→A	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600404-A-T
50	Q→L	missense_variant	gnomAD	—	6.84e-07	damaging	likely_benign (0.31)	-9.19	chr5-139600406-A-T
51	C→C	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600410-T-C
52	S→S	synonymous_variant	gnomAD	—	2.05e-06	—	—	0.00	chr5-139600413-A-T
53	A→S	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.78)	-9.19	COSV108064706
55	P→P	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600422-T-C
58	D→D	synonymous_variant	gnomAD	—	4.10e-06	—	—	0.00	chr5-139600431-T-C
58	D→E	missense_variant	gnomAD	—	2.05e-06	—	likely_benign (0.17)	-3.06	chr5-139600431-T-G
59	D→D	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139600434-T-C
59	D→N	missense_variant	COSMIC	—	—	damaging	ambiguous (0.56)	-8.00	COSV54079619
59	D→D	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV54080651
60	M→T	missense_variant	gnomAD	—	6.85e-07	damaging	ambiguous (0.47)	-8.18	chr5-139614586-T-C
61	F→F	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV99551398
61	F→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-8.62	COSV54080918
61	F→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-8.62	COSV108064633
62	H→H	synonymous_variant	gnomAD	—	6.85e-07	—	—	0.00	chr5-139614593-T-C
65	A→V	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.98)	-9.94	COSV54080672
66	T→I	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.94)	-10.31	COSV54080777
67	I→V	missense_variant	gnomAD	—	6.84e-07	—	likely_benign (0.25)	-6.97	chr5-139614606-A-G
67	I→V	missense_variant	ClinVar	Uncertain significance	—	—	likely_benign (0.25)	-6.97	ClinVar:2331085
70	P→S	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.92)	-8.37	COSV54077336
72	D→D	synonymous_variant	gnomAD	—	6.71e-05	—	—	0.00	chr5-139614702-C-T
73	S→N	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.92)	-9.43	chr5-139614704-G-A
74	P→A	missense_variant	gnomAD	—	2.05e-06	—	ambiguous (0.37)	-6.47	chr5-139614706-C-G
74	P→L	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.83)	-8.31	chr5-139614707-C-T
76	Q→E	missense_variant	COSMIC	—	—	—	likely_benign (0.11)	-7.27	COSV54077632
77	G→S	missense_variant	gnomAD	—	6.84e-07	—	likely_benign (0.17)	-3.15	chr5-139614715-G-A
78	G→G	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614720-A-G
82	L→L	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614730-T-C
83	T→I	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.74)	-7.86	chr5-139614734-C-T
85	H→R	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.68)	-8.25	COSV54080726
86	F→F	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV107280747
87	P→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.97)	-9.50	COSV108064632
91	P→P	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614759-C-A
91	P→P	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614759-C-T
91	P→S	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-10.94	COSV54077205
93	K→R	missense_variant	gnomAD	—	6.84e-07	damaging	likely_benign (0.28)	-8.75	chr5-139614764-A-G
93	K→R	missense_variant	COSMIC	—	—	damaging	likely_benign (0.28)	-8.75	COSV54078386
94	P→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.98)	-10.00	COSV105026663
97	V→V	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614863-T-C
100	T→K	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.70)	-9.87	COSV54080505
101	T→T	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614875-A-T
102	R→R	synonymous_variant	gnomAD	—	2.05e-06	—	—	0.00	chr5-139614876-A-C
102	R→I	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.97)	-11.93	COSV54077275
104	Y→C	missense_variant	gnomAD	—	1.37e-06	damaging	likely_pathogenic (0.86)	-9.56	chr5-139614883-A-G
104	Y→Y	synonymous_variant	gnomAD	—	1.37e-06	—	—	0.00	chr5-139614884-T-C
104	Y→Y	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV105026661
105	H→H	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614887-T-C
105	H→R	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-9.56	COSV99551272
105	H→N	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-8.62	COSV99551494
105	H→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.98)	-10.56	COSV99551495
106	P→P	synonymous_variant	gnomAD	—	9.92e-05	—	—	0.00	chr5-139614890-A-G
107	N→N	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614893-T-C
111	N→N	synonymous_variant	gnomAD	—	6.16e-06	—	—	0.00	chr5-139614905-T-C
112	G→S	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-8.87	COSV54080718
112	G→D	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-10.19	COSV54077127
112	G→G	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV99551303
113	S→N	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.74)	-8.12	chr5-139614910-G-A
113	S→S	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV54079119
119	L→L	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139614927-C-T
119	L→L	synonymous_variant	gnomAD	—	8.89e-06	—	—	0.00	chr5-139614929-A-G
120	R→R	synonymous_variant	gnomAD	—	7.53e-06	—	—	0.00	chr5-139614932-A-G
120	R→*	stop_gained	COSMIC	—	—	LoF	—	—	COSV54077789
120	R→Q	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.88)	-9.87	COSV54081074
121	S→P	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.91)	-11.00	chr5-139614933-T-C
121	S→S	synonymous_variant	gnomAD	—	1.37e-06	—	—	0.00	chr5-139614935-A-T
122	Q→Q	synonymous_variant	gnomAD	—	1.37e-06	—	—	0.00	chr5-139614938-G-A
122	Q→H	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.92)	-7.87	COSV99551304
123	W→C	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-9.81	COSV54081180
124	S→S	synonymous_variant	gnomAD	—	1.37e-06	—	—	0.00	chr5-139614944-T-C
124	S→F	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-11.81	COSV108786173
127	L→L	synonymous_variant	gnomAD	—	6.85e-07	—	—	0.00	chr5-139614951-C-T
127	—	frameshift_variant	COSMIC	—	—	LoF	—	—	COSV54078283
128	T→I	missense_variant	gnomAD	—	6.85e-07	damaging	likely_pathogenic (1.00)	-10.62	chr5-139614955-C-T
128	T→I	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (1.00)	-10.62	COSV54079667
130	S→L	missense_variant	gnomAD	—	6.85e-07	damaging	likely_pathogenic (0.96)	-9.62	chr5-139614961-C-T
130	S→S	synonymous_variant	gnomAD	—	3.43e-06	—	—	0.00	chr5-139614962-A-C
130	S→S	synonymous_variant	gnomAD	—	6.85e-07	—	—	0.00	chr5-139614962-A-T
134	L→L	synonymous_variant	gnomAD	—	2.09e-06	—	—	0.00	chr5-139623373-T-C
135	—	frameshift_variant	gnomAD	—	6.97e-07	LoF	—	—	chr5-139623377-CCATCTGTT-C
137	C→C	synonymous_variant	gnomAD	—	6.90e-07	—	—	0.00	chr5-139623384-T-C
138	S→T	missense_variant	gnomAD	—	6.90e-07	damaging	likely_pathogenic (0.91)	-11.44	chr5-139623385-T-A
138	S→F	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-10.75	COSV105026656
139	L→L	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623390-G-C
139	L→L	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV54077579
140	L→L	synonymous_variant	gnomAD	—	8.91e-03	—	—	0.00	chr5-139623391-T-C
140	L→L	synonymous_variant	COSMIC	—	—	—	—	0.00	COSV54079502
141	C→C	synonymous_variant	gnomAD	—	6.88e-07	—	—	0.00	chr5-139623396-T-C
143	P→A	missense_variant	gnomAD	—	6.88e-07	damaging	likely_pathogenic (0.61)	-4.52	chr5-139623400-C-G
143	P→P	synonymous_variant	gnomAD	—	6.88e-07	—	—	0.00	chr5-139623402-C-G
144	N→D	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.98)	-10.12	COSV54077287
146	D→V	missense_variant	gnomAD	—	6.88e-07	damaging	likely_pathogenic (0.95)	-11.31	chr5-139623410-A-T
146	D→D	synonymous_variant	gnomAD	—	6.88e-07	—	—	0.00	chr5-139623411-T-C
147	D→Y	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.99)	-10.12	COSV99551400
148	P→P	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623417-T-C
148	P→P	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623417-T-G
149	L→L	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623420-A-G
149	L→*	stop_gained	COSMIC	—	—	LoF	—	—	COSV99551309
151	P→T	missense_variant	gnomAD	—	6.89e-07	damaging	likely_pathogenic (0.92)	-8.19	chr5-139623424-C-A
151	P→P	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623426-T-A
151	P→L	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.93)	-8.00	COSV54077732
152	E→K	missense_variant	gnomAD	—	6.89e-07	damaging	likely_pathogenic (0.95)	-10.05	chr5-139623427-G-A
152	E→E	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623429-G-A
153	I→M	missense_variant	gnomAD	—	6.89e-07	damaging	likely_pathogenic (0.72)	-8.19	chr5-139623432-T-G
153	I→V	missense_variant	ClinVar	Uncertain significance	—	—	ambiguous (0.37)	-6.94	ClinVar:3976716
154	A→T	missense_variant	gnomAD	—	6.89e-07	damaging	likely_pathogenic (0.98)	-8.75	chr5-139623433-G-A
154	A→G	missense_variant	gnomAD	—	6.89e-07	—	ambiguous (0.45)	-7.15	chr5-139623434-C-G
154	A→A	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623435-T-C
155	R→R	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623436-C-A
155	R→G	missense_variant	gnomAD	—	6.89e-07	damaging	likely_pathogenic (0.60)	-7.87	chr5-139623436-C-G
155	R→Q	missense_variant	gnomAD	—	3.45e-06	—	ambiguous (0.34)	-6.62	chr5-139623437-G-A
155	R→Q	missense_variant	COSMIC	—	—	—	ambiguous (0.34)	-6.62	COSV54080979
156	I→L	missense_variant	gnomAD	—	6.88e-07	—	likely_benign (0.14)	-5.20	chr5-139623439-A-C
157	Y→C	missense_variant	gnomAD	—	6.89e-07	—	likely_benign (0.33)	-6.65	chr5-139623443-A-G
159	T→T	synonymous_variant	gnomAD	—	6.89e-07	—	—	0.00	chr5-139623450-A-C
159	T→I	missense_variant	COSMIC	—	—	—	ambiguous (0.50)	-6.29	COSV107280745
161	R→G	missense_variant	gnomAD	—	6.90e-07	—	ambiguous (0.48)	-6.93	chr5-139623454-A-G
161	R→T	missense_variant	gnomAD	—	6.90e-07	—	ambiguous (0.43)	-6.49	chr5-139623455-G-C
162	E→Q	missense_variant	gnomAD	—	6.90e-07	damaging	likely_benign (0.21)	-7.67	chr5-139623457-G-C
162	E→K	missense_variant	COSMIC	—	—	damaging	ambiguous (0.39)	-8.05	COSV54077797
164	Y→Y	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139626759-C-T
166	R→G	missense_variant	gnomAD	—	1.37e-06	—	likely_benign (0.21)	-7.18	chr5-139626763-A-G
167	I→V	missense_variant	gnomAD	—	6.84e-07	—	likely_benign (0.09)	-3.55	chr5-139626766-A-G
168	A→V	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.87)	-5.52	COSV54079876
169	R→W	missense_variant	gnomAD	—	2.05e-06	damaging	ambiguous (0.49)	-8.05	chr5-139626772-C-T
169	R→Q	missense_variant	gnomAD	—	1.37e-06	—	likely_benign (0.10)	-2.28	chr5-139626773-G-A
169	R→L	missense_variant	gnomAD	—	6.84e-07	—	ambiguous (0.37)	-6.48	chr5-139626773-G-T
169	R→P	missense_variant	ClinVar	Uncertain significance	—	damaging	likely_pathogenic (0.99)	-9.92	ClinVar:3330554
169	R→Q	missense_variant	COSMIC	—	—	—	likely_benign (0.10)	-2.28	COSV99551584
170	E→K	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.77)	-8.75	chr5-139626775-G-A
170	E→K	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.77)	-8.75	COSV54080517
172	T→A	missense_variant	COSMIC	—	—	damaging	likely_pathogenic (0.66)	-6.23	COSV54077550
175	Y→C	missense_variant	gnomAD	—	6.84e-07	damaging	likely_pathogenic (0.88)	-8.25	chr5-139626791-A-G
175	Y→F	missense_variant	gnomAD	—	1.37e-06	—	likely_benign (0.20)	-5.93	chr5-139626791-A-T
175	Y→Y	synonymous_variant	gnomAD	—	6.84e-07	—	—	0.00	chr5-139626792-T-C
176	A→A	synonymous_variant	gnomAD	—	1.10e-05	—	—	0.00	chr5-139626795-G-A
177	M→T	missense_variant	gnomAD	—	6.85e-07	—	ambiguous (0.51)	-7.30	chr5-139626797-T-C
177	M→T	missense_variant	ClinVar	Uncertain significance	—	—	ambiguous (0.51)	-7.30	ClinVar:4601227

167 variants in the shared canonical core.

LoF = frameshift / stop-gain / splice-disrupting — inherently loss-of-function, flagged by consequence (AlphaMissense and ESM-C score only missense/substitutions, so they are blank here by design, not by absence of impact). AlphaMissense is computed in the canonical reading frame, so it scores the shared core but reads N/A across an isoform-unique extension — use ESM-C ΔLLR there.

UBE2D2 · ENST00000398733.8

Folding

Evidence — click any tile for the differential-region detail

Comparison

Comparison

Comparison

Comparison

Comparison

Comparison

Details

Comparison

Comparison

Comparison

Comparison

Comparison

Comparison

Comparison

Elements and coordinates

Below threshold

Comparison

Details

Comparison

Part 1 · Global feature comparison

Isoform-only features — 153 total (prevalence ≥ 2)

Canonical-only features — 34 total (prevalence ≥ 2)

Shared features by |Δ| activation — 916 total (prevalence ≥ 2)

Part 2 · Differential coordinates

Unique-region features — 241 distinct (prevalence ≥ 2)

Clinical variants

Differential region — N-terminal extension (isoform-unique)

Shared canonical core — sequence common to canonical and isoform; AlphaMissense applies here