CATH FunFam Annotation v5 — One concept per slide

🧬 CATH FunFam Annotation

Turn a FunFam
into a description

Input = FunFam ID · Output = description
If the FunFam mixes multiple functions, emit one description per subgroup

1

Input — what a CATH FunFam member looks like

Input

You give us

FunFam ID

3.40.50.620/FF/000152

short alias: FF152

We parse the Stockholm (.sto) file

# raw CATH FunFam alignment file
#=GS P41972/11-192_400-630 OS Staphylococcus aureus
#=GS P41972/11-192_400-630 DE Isoleucine--tRNA ligase
#=GS P41972/11-192_400-630 DR EC; 6.1.1.5
P41972/11-192_400-630  --MDYKETLLMPKTDFPMRGG...
#=GS Q2FZ82/18-243 OS Staphylococcus aureus
#=GS Q2FZ82/18-243 DE Isoleucine--tRNA ligase
#=GS Q2FZ82/18-243 DR EC; 6.1.1.5
Q2FZ82/18-243          --MDYKETLNLPKTSFPMRGD...
...

→ Parsed into a list of members (5 of 43 shown)

UniProt AC	cath_id (residue range)	organism	description	EC
P41972	1qu2A01/11-192_400-630	Staphylococcus aureus	Isoleucine--tRNA ligase	6.1.1.5
Q2FZ82	Q2FZ82/18-243	Staphylococcus aureus	Isoleucine--tRNA ligase	6.1.1.5
Q8RH47	Q8RH47/15-230	Fusobacterium nucleatum	Isoleucine--tRNA ligase	6.1.1.5
G4FF97	G4FF97/18-269	(missing in UniProt)	—	6.1.1.5
P46213	P46213/18-269	Thermotoga maritima	Isoleucine--tRNA ligase	6.1.1.5

This demo

FF152: 45 rows / 43 unique ACs · FF153: 46 rows / 42 unique ACs · total = 91 rows → 85 distinct proteins

Note: some ACs appear multiple times (multi-domain proteins). We deduplicate by AC.

So far we have …

… only 5 raw fields per member (cath_id, AC, range, description, EC) — and CATH's own description / EC may be stale.
Next: enrich each AC from 7 more public sources.

2

Glossary — abbreviations used in this report

Reference

Every abbreviation that appears later in the report is defined here. Skim or skip — come back if something is unclear.

CATH / databases

CATH	Class · Architecture · Topology · Homology — protein structure classification.
FunFam	Functional Family — a sub-grouping inside a CATH superfamily. Each FunFam has its own multiple-sequence alignment.
FF152 / FF153	Short alias for FunFam 000152 / 000153 in the demo. Full ID: 3.40.50.620/FF/000152.
cath_id	CATH's identifier for one domain instance, e.g. 1qu2A01/11-192 = PDB 1qu2 chain A, residues 11–192.
AC	UniProt Accession Code, e.g. P41972.
Stockholm	File format for multiple-sequence alignments (.sto). What CATH publishes for each FunFam.
Pfam / InterPro / KEGG / OrthoDB	External cross-reference databases. UniProt links to all of them.
REST	REpresentational State Transfer — UniProt's web API style.
IUBMB	International Union of Biochemistry and Molecular Biology — the body that maintains the EC numbering scheme.

Annotation fields

EC	Enzyme Commission number — a 4-part identifier of the catalysed reaction (e.g. 6.1.1.5).
GO	Gene Ontology — controlled-vocabulary terms about gene function.
MF / BP / CC	GO sub-ontologies: Molecular Function / Biological Process / Cellular Component.
function_text	Free-text functional description from UniProt's FUNCTION field.
catalytic_activity	UniProt's CATALYTIC ACTIVITY field — the reaction equation.

Structure metrics

pLDDT	predicted Local Distance Difference Test — AlphaFold's per-residue confidence (0–100). ≥ 90 = very high.
PAE	Predicted Aligned Error — AlphaFold's predicted positional error matrix.
MSA	Multiple Sequence Alignment.
TM-score	Template Modelling score — Foldseek/Foldcompare global similarity, 0–1. > 0.5 = same fold. > 0.8 = very similar.
lddt	Local Distance Difference Test — Foldseek per-residue structural quality.
fident	Fraction identical — sequence identity inside the structural alignment.
GDT-TS / GDT-HA	Global Distance Test, Total Score / High Accuracy variants. Folddisco geometry score.
chamfer / hausdorff	Two distance metrics between residue sets. Folddisco geometry scores.
RMSD	Root Mean Squared Deviation between atoms (Angstrom).
idf	Inverse Document Frequency — borrowed from text retrieval. In Folddisco, rarer motif features score higher idf, so a high idf = more specific / unique match.

Biology specific to this demo

IleRS	Ileucyl-tRNA Synthetase — attaches Ile to tRNA(Ile). EC 6.1.1.5. The protein FF152 is named after.
KMSKS / HIGH-motif	Two conserved sequence motifs in Class-I aminoacyl-tRNA synthetases (including IleRS). Part of the ATP-binding site.
CP1 domain	Connective Polypeptide 1 — an editing sub-domain in IleRS that hydrolyses mis-activated Val-AMP.
APR	Adenosine-5'-Phosphosulfate Reductase. Uses APS as substrate. EC 1.8.4.9. FF153 majority.
PAPR	Phospho-Adenosine 5'-Phosphosulfate Reductase. Uses PAPS. EC 1.8.4.10. FF153 minority.
APS / PAPS	Substrates: Adenosine 5'-Phosphosulfate / 3'-Phospho-Adenosine 5'-Phosphosulfate.
GSH / Trx	Electron donors: Glutathione (used by APR) / Thioredoxin (used by PAPR).
[4Fe-4S]	Iron-sulfur cluster cofactor — 4 Fe + 4 S held by 4 cysteines. PAPR has one; helps shuttle electrons.
AMP / ATP / DTT	Adenosine mono- / tri-phosphate / dithiothreitol (a lab reductant sometimes used instead of GSH/Trx).
PURE / SPLIT	Our verdict labels: PURE = one description suffices; SPLIT = the FunFam contains multiple functions, emit N descriptions.
majority / minority	After k=2 clustering, the larger group is "majority", smaller is "minority".

3

The Problem — is a FunFam really functionally pure?

Problem

CATH hands you a FunFam ID containing N proteins.
How do you know they really do the same thing?
Or are two or more functions quietly mixed in?

📥

Input

FunFam ID

e.g. FF152

→

🤔

Question

Pure or Mixed?

this is the hard part

→

📝

Output

Description

pure → 1
mixed → N

4

Our 3-step pipeline

Approach

Don't trust CATH's own annotation (may be stale). For each protein, re-fetch from multiple public sources, then judge by consensus.

🔍

Step 1

Collect

For each UniProt AC, pull data from
8 sources →
one per-member dict

→

⚖️

Step 2

Compare

Pairwise
sequence / structure / GO
similarity → look for clusters

→

⚖️

Step 3

Decide

k=2 clustering +
discriminator (EC/kingdom)
→ PURE / SPLIT

Next 5 slides: what each of Step 1's 8 sources looks like (screenshots + real output).

5

Source ① — UniProt REST

Step 1 / 8

https://www.uniprot.org/uniprotkb/P41972/entry · screenshot from UniProt web page

API endpoint

GET https://rest.uniprot.org/uniprotkb/
     P41972.json

Fields we extract (30+)

protein_nameorganism lineagekingdom ecgo_mfgo_bpgo_cc function_textcatalytic_activity cofactoractive_sitebinding_site keywordspdb_xrefs InterProPfamReactome KEGGPubMed refs

UniProt is the ground truth: protein name, organism, EC, GO, catalytic activity all come from here.
CATH's own EC annotation can lag behind (we see a real case on slide 17).

6

Source ② — AlphaFold DB

Step 1 / 8

GET alphafold.ebi.ac.uk/api/prediction/P41972

[{
  "modelEntityId": "AF-P41972-F1",
  "globalMetricValue": 95.56,
  "fractionPlddtVeryHigh": 0.908,
  "fractionPlddtConfident": 0.091,
  "fractionPlddtLow": 0.001,
  "fractionPlddtVeryLow": 0.0,
  "latestVersion": 6,
  "pdbUrl": ".../AF-P41972-F1-model_v6.pdb",
  "paeImageUrl": ".../predicted_aligned_error_v6.png",
  "msaUrl": ".../AF-P41972-F1-msa_v6.a3m",
  "gene": "ileS",
  "organismScientificName": "Staphylococcus aureus"
}]

Extracted

pLDDT_mean frac_very_high frac_confident frac_low model_version model_url

Available but not used yet

PAE matrix MSA file mmCIF

pLDDT = AlphaFold's per-residue confidence (0–100).
Mean ≥ 90 = the whole structure is trustworthy; < 50 = that region is disordered.

We also download the PDB to structures/P41972.pdb for use by Foldseek / Folddisco below.

7

Source ③ — Foldseek (structural all-vs-all)

Step 1 / 8

Local tool. Pairwise structural similarity over all AF models. Returns 17 columns; we currently use 1.

foldseek easy-search structures/P41972.pdb structures/Q5ZKA2.pdb out tmp \
  --format-output query,target,fident,alnlen,mismatch,gapopen,
  qstart,qend,tstart,tend,evalue,bits,alntmscore,lddt,prob,qlen,tlen

Real output (P41972 vs Q5ZKA2)

query	target	fident	alnlen	mismatch	gapopen	qstart	qend	tstart	tend	evalue	bits	alntmscore	lddt	prob	qlen	tlen
P41972	Q5ZKA2	0.360	979	539	29	1	916	44	997	1.59e-80	3379	0.8613	0.8011	1.000	917	1000

Used

alntmscore = 0.8613 — alignment-normalized TM-score, quantifies overall structural similarity.
1.0 = identical, > 0.5 = same fold.

Not yet used (could add)

lddt = 0.801(local structure quality)
fident = 0.360(seq identity within the structural alignment)
alnlen / qlen / tlen(coverage)

8

Source ④ — Folddisco (active-site motif search)

Step 1 / 8

Local tool. Given a few active-site residues of a reference protein, look across the FunFam for the same geometric motif.

folddisco query -p structures/O05927.pdb \
  -q 'A139,A140,A228,A231,A256'  # PAPR [4Fe-4S] cluster + catalytic Cys
  -i funfam_index --format-output tid,nid,idf,rmsd,tm_score,
  gdt_ts,gdt_ha,chamfer_distance,hausdorff_distance,
  matching_residues,query_residues

Real output (top 3 hits)

tid	idf	rmsd	tm_score	gdt_ts	gdt_ha	chamfer	hausdorff	matching_residues	query_residues
O05927 (self)	30.00	0.000	1.000	1.000	1.000	0.000	0.000	A139,A140,A228,A231,A256	A139,...
P92981 (APR)	10.00	0.500	0.396	1.000	0.775	0.443	0.886	A194,A195,A286,A289,A314	A139,...
P41972 (IleRS — unexpected cross-hit; serves as negative control)	0.00	0.360	0.458	1.000	0.875	0.330	0.545	_,_,A906,A909,_	A139,...

Used

matched_residues (residue hit count), idf (score), rmsd.
Scores whether the motif exists in this member.

Not yet used (high-information)

tm_score / gdt_ts / gdt_ha three geometry-similarity scores
chamfer / hausdorff two distance metrics
matching_residues most important: tells you where query A139 maps in the target (e.g. P92981 A194)

9

Sources ⑤⑥⑦⑧ — literature + Stockholm + derived

Step 1 / 8

⑤ EuropePMC — literature

europepmc.org/search?query=P41972 — 4 hits

GET ebi.ac.uk/europepmc/webservices/rest/search?query=P41972&format=json
Extract: pmid title journal year authors · top 5 papers per protein

⑥ CATH Stockholm — raw annotation

# local .sto parsing
#=GS P41972/11-192_400-630 OS Staphylococcus aureus
#=GS P41972/11-192_400-630 DE Isoleucine--tRNA ligase
#=GS P41972/11-192_400-630 DR EC; 6.1.1.5
P41972/11-192_400-630   --MDYKETLLMPKTDFPMRGG...

Extract: cath_id organism description ec aligned_sequence
The EC here can be stale (an example on slide 17).

⑦ MSA % identity (derived)

Pairwise identity over ⑥'s Stockholm alignment:
P41972 vs Q5ZKA2 → 0.504
P41972 vs Q8RH47 → 0.819
Gap-aware: counts only columns where both sides are non-gap.

⑧ GO Jaccard (derived)

Pairwise Jaccard over ① UniProt's GO sets:
|GO_A ∩ GO_B| / |GO_A ∪ GO_B|
Missing GO → cell left empty, excluded from the ensemble.

10

After collecting all 8 sources, each protein looks like this

Per-member dict

Each colored block = number of fields one source gives for P41972. Slide 18 lets you click any AC interactively.

5

📚

CATH Stockholm

cath_id, name, EC, organism, aligned_seq

30+

🧬

UniProt REST

name, EC, GO MF/BP/CC, function_text, catalytic_activity, cofactor, binding_site, PDB xrefs, InterPro, Pfam, …

7

🔮

AlphaFold DB

pLDDT_mean (95.6), frac_very_high (0.908), model_url, version

N×1

📐

Foldseek

TM-score vector vs the other N FF members

2

🎯

Folddisco

matched_residues / score / rmsd for 2 reference motifs

≤5

📄

EuropePMC

top 5 publications · pmid / title / journal / year

N×1

📊

MSA % identity

identity vector vs the other N FF members

N×1

🔗

GO Jaccard

function-descriptor similarity vs the other N FF members

      This is the output of Step 1 — every UniProt AC gets a full dict. All 85 are written to member_facts.json + the Excel.
    

11

Step 2 — three similarity layers find sub-groups

Compare

For each protein, compute pairwise sequence / structure / GO similarity vs every other FF member → N×N matrix. If all three matrices show the same 2-block pattern, the sub-groups are real, not noise.

FF152

FF153

12

Step 3 — from similarity to a PURE / SPLIT verdict

Decide

Four steps: ① coverage-weighted fusion of the 3 similarities → ② hierarchical clustering into 2 groups → ③ compute EC/kingdom/name purity → ④ any ≥ 0.8 → SPLIT.

① Ensemble

Weighted fusion
sim = Σ wₖ sₖ / Σ wₖ
w_MSA=.30, w_Fs=.40, w_GO=.30

→

② Cluster k=2

scipy
average linkage
split into 2

→

③ Discriminator

EC > kingdom > name
purity ≥ 0.8 ?
→ SPLIT (and which axis)

Worked example

FF	cluster	n	EC purity	Kingdom purity	verdict
FF152	majority	39	6.1.1.5 = 100%	Bacteria = 100%	SPLIT (kingdom)
FF152	minority	6	6.1.1.5 = 100%	Eukaryota = 100%	→ eukaryotic mitochondrial IleRS
FF153	majority	28	1.8.4.9 = 100%	Eukaryota = 100%	SPLIT (EC)
FF153	minority	18	1.8.4.10 = 100%	Bacteria = 100%	→ bacterial PAPR

13

Result ① — FF152: SPLIT

SPLIT

      FF152 is not pure. CATH's name "Isoleucine--tRNA ligase" is correct, but inside are 39 bacterial + 6 eukaryotic mitochondrial IleRS — two different sub-cellular systems.
    

🦠

majority

Bacterial cytoplasmic IleRS

n = 37

Staphylococcus / Fusobacterium / Thermotoga / etc.

🧬

minority

Eukaryotic mitochondrial IleRS

n = 6

Minority members (all 6)

D7EHQ7Tribolium castaneum

F0Z9P8Dictyostelium purpureum

Q5ZKA2Gallus gallus

G1K9T7Anolis carolinensis

U3I8X9Anas platyrhynchos platyrhynchos

F1P399—

Why it splits

Kingdom purity = 100% on both sides (Bacteria vs Eukaryota); every minority protein name contains "mitochondrial". EC is the same (6.1.1.5) — same reaction, but completely different localization, lineage, and N-terminal targeting peptide.

14

FF152 — two descriptions

Output

FF152.a — Bacterial cytoplasmic IleRS (majority)

Isoleucine--tRNA ligase · EC 6.1.1.5 · Bacteria · n=39 (0 Swiss-Prot) · cofactor: Zn(2+) · localization: Cytoplasm

Function: Catalyzes the attachment of isoleucine to tRNA(Ile). As IleRS can inadvertently accommodate and process structurally similar amino acids such as valine, to avoid such errors it has two additional distinct tRNA(Ile)-dependent editing activities. One activity is designated as 'pretransfer' editing and involves the hydrolysis of activated Val-AMP. The other activity is designated 'posttransfer' editing and involves deacylation of mischarged Val-tRNA(Ile) (By similarity)

Catalytic activity: tRNA(Ile) + L-isoleucine + ATP = L-isoleucyl-tRNA(Ile) + AMP + diphosphate

Known PDB: 1FFY, 1QU2, 1QU3

References: PMID 8163160: Analysis and toxic overexpression in Escherichia coli of a staphylococcal gene encoding is; PMID 10446055: Insights into editing from an Ile-tRNA synthetase structure with tRNAIle and mupirocin.

FF152.b — Eukaryotic mitochondrial IleRS (minority)

isoleucine--tRNA ligase · EC 6.1.1.5 · Eukaryota · n=6 (0 Swiss-Prot) · localization: Mitochondrion

Function: Aminoacyl-tRNA synthetase that catalyzes the specific attachment of isoleucine to its cognate tRNA (tRNA(Ile))

Catalytic activity: tRNA(Ile) + L-isoleucine + ATP = L-isoleucyl-tRNA(Ile) + AMP + diphosphate

References: PMID 18362917: The genome of the model beetle and pest Tribolium castaneum.; PMID 19820115: BeetleBase in 2010: revisions to provide comprehensive genomic information for Tribolium c

15

Result ② — FF153: SPLIT

SPLIT

      FF153 is not pure. CATH's name "Phosphoadenosine phosphosulfate reductase" describes only the bacterial branch (PAPR); the family also mixes in a plant / cyanobacterial APR branch (uses APS, not PAPS; GSH, not Trx).
    

🌱

majority (n=28)

APR (GSH-dep)

EC 1.8.4.9

eukaryota + cyanobacteria · APS substrate

🦠

minority (n=14)

PAPR (Trx-dep)

EC 1.8.4.10

bacteria · PAPS substrate

Minority members (first 12)

O05927Pseudomonas aeruginosa

A0A157WSG9—

A0A335M5J3—

A0A1F0I098—

A0A1S1C0B4—

A0A3S4MQM5—

W1MKF2—

A0A0C7CX40—

A0A1C7BU64—

Q02KP7Pseudomonas aeruginosa

B7VBC3Pseudomonas aeruginosa

A0A335EVG0—

… +2 more(other bacterial PAPR)

Why it splits

Both subgroups show 100% EC purity (1.8.4.9 vs 1.8.4.10). Different substrates (APS vs PAPS), different electron donors (GSH vs Trx) — these are two different biochemical reactions, not variants.

16

FF153 — two descriptions

Output

FF153.a — APR / GSH-dependent (majority)

adenylyl-sulfate reductase (glutathione) · EC 1.8.4.9 · Eukaryota · n=28 (0 Swiss-Prot) · cofactor: [4Fe-4S] cluster · localization: Chloroplast, Plastid

Function: Reduces sulfate for Cys biosynthesis. Substrate preference is adenosine-5'-phosphosulfate (APS) >> 3'-phosphoadenosine-5'-phosphosulfate (PAPS). Uses glutathione or DTT as source of protons

Catalytic activity: glutathione disulfide + sulfite + AMP + 2 H(+) = adenosine 5'-phosphosulfate + 2 glutathione

Known PDB: 5YRY

References: PMID 8917599: Three members of a novel small gene-family from Arabidopsis thaliana able to complement fu; PMID 11130712: Sequence and analysis of chromosome 1 of the plant Arabidopsis thaliana.

FF153.b — PAPR / Trx-dependent (minority)

Adenosine 5'-phosphosulfate reductase · EC 1.8.4.10 · Bacteria · n=18 (0 Swiss-Prot) · cofactor: [4Fe-4S] cluster · localization: Cytoplasm

Function: Catalyzes the formation of sulfite from adenosine 5'-phosphosulfate (APS) using thioredoxin as an electron donor

Catalytic activity: [thioredoxin]-disulfide + sulfite + AMP + 2 H(+) = adenosine 5'-phosphosulfate + [thioredoxin]-dithiol

Known PDB: 2GOY

References: PMID 9218775: Cloning, sequence and mutagenesis of the structural gene of Pseudomonas aeruginosa CysB, w; PMID 10984043: Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen.

17

Side finding — CATH annotation is stale

Bonus

Issue

FF153 Stockholm labels O05927 as EC 1.8.4.8, but the current UniProt assignment is EC 1.8.4.10.

Why

1.8.4.8 (PAPS reductase, uses PAPS) is still valid. As enzymology resolved the true substrate, APS reductases got their own numbers — 1.8.4.9 (glutathione, 2000) and 1.8.4.10 (thioredoxin, 2003). Bacterial members filed as 1.8.4.8 actually use APS and were reassigned to 1.8.4.10; CATH kept the old number.

How we caught it

Our dict keeps ec_from_stockholm and UniProt's ec separately. Any disagreement is a stale signal.

18

Explore + download

Try it

Click any AC for its full dict

AC	FF	organism	protein name	EC
P41972	FF152	Staphylococcus aureus	Isoleucine--tRNA ligase	6.1.1.5
Q2FZ82	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q8RH47	FF152	Fusobacterium nucleatum subsp.	Isoleucine--tRNA ligase	6.1.1.5
G4FF97	FF152	—	—	—
P46213	FF152	Thermotoga maritima (strain AT	Isoleucine--tRNA ligase	6.1.1.5
P67509	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
D7EHQ7	FF152	Tribolium castaneum	isoleucine--tRNA ligase	6.1.1.5
F0Z9P8	FF152	Dictyostelium purpureum	isoleucine--tRNA ligase	6.1.1.5
Q5ZKA2	FF152	Gallus gallus	Isoleucine--tRNA ligase, mitochond	6.1.1.5
A0A0Y1XXB4	FF152	—	—	—
A0A2S4EPE4	FF152	—	—	—
G1K9T7	FF152	Anolis carolinensis	isoleucine--tRNA ligase	6.1.1.5
A5IML2	FF152	Thermotoga petrophila (strain	Isoleucine--tRNA ligase	6.1.1.5
A0A1J4HCQ3	FF152	—	—	—
Q5HPZ9	FF152	Staphylococcus epidermidis (st	Isoleucine--tRNA ligase	6.1.1.5
A0A1E8WR39	FF152	—	—	—
A0A2S6DX65	FF152	—	—	—
A0A2N5RN72	FF152	—	—	—
A0A380H4S2	FF152	Staphylococcus saccharolyticus	Isoleucine--tRNA ligase	6.1.1.5
A0A432A8Z4	FF152	—	—	—
F9EMC2	FF152	Fusobacterium animalis ATCC 51	Isoleucine--tRNA ligase	6.1.1.5
U3I8X9	FF152	Anas platyrhynchos platyrhynch	isoleucine--tRNA ligase	6.1.1.5
F1P399	FF152	—	—	—
D5RE11	FF152	—	—	—
Q2YXH4	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q8NX29	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q6GA19	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q846V6	FF152	Staphylococcus epidermidis	Isoleucine--tRNA ligase	6.1.1.5
A0A0E1AHI6	FF152	—	—	—
A0A0E1VIB0	FF152	—	—	—
A0A0D6W8U9	FF152	—	—	—
P67508	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
A6QG93	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q5HGN8	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
A5IS79	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q2FHP4	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
A6U113	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
A7X1D8	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
A8Z3N1	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q8CSX1	FF152	Staphylococcus epidermidis (st	Isoleucine--tRNA ligase	6.1.1.5
A0A482QXK9	FF152	—	—	—
Q6GHP2	FF152	Staphylococcus aureus (strain	Isoleucine--tRNA ligase	6.1.1.5
Q4L5P4	FF152	Staphylococcus haemolyticus (s	Isoleucine--tRNA ligase	6.1.1.5
O05927	FF153	Pseudomonas aeruginosa (strain	Adenosine 5'-phosphosulfate reduct	1.8.4.10
P92981	FF153	Arabidopsis thaliana	5'-adenylylsulfate reductase 2, ch	1.8.4.9
P92979	FF153	Arabidopsis thaliana	5'-adenylylsulfate reductase 1, ch	1.8.4.9
P92980	FF153	Arabidopsis thaliana	5'-adenylylsulfate reductase 3, ch	1.8.4.9
F4HX50	FF153	Arabidopsis thaliana	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A157WSG9	FF153	—	—	—
A0A335M5J3	FF153	—	—	—
A0A072VFH5	FF153	Medicago truncatula	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A1F0I098	FF153	—	—	—
A0A1S1C0B4	FF153	—	—	—
A0A3S4MQM5	FF153	—	—	—
V4JUZ5	FF153	Eutrema salsugineum	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A078GHM5	FF153	Brassica napus	adenylyl-sulfate reductase (glutat	1.8.4.9
R0IAW2	FF153	Capsella rubella	adenylyl-sulfate reductase (glutat	1.8.4.9
D7KUT5	FF153	Arabidopsis lyrata subsp. lyra	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A3P5Z8Q6	FF153	—	—	—
A0A3P6C6V3	FF153	—	—	—
M4F073	FF153	Brassica campestris	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A0D3BVL7	FF153	Brassica oleracea var. olerace	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A072V4H5	FF153	Medicago truncatula	adenylyl-sulfate reductase (glutat	1.8.4.9
R0FF83	FF153	Capsella rubella	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A0D3E884	FF153	Brassica oleracea var. olerace	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A3P6DHS6	FF153	—	—	—
A0A0D3DFC9	FF153	Brassica oleracea var. olerace	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A3P6E7E2	FF153	—	—	—
D7M1S0	FF153	Arabidopsis lyrata subsp. lyra	adenylyl-sulfate reductase (glutat	1.8.4.9
M4DSB3	FF153	Brassica campestris	adenylyl-sulfate reductase (glutat	1.8.4.9
V4L3C0	FF153	Eutrema salsugineum	adenylyl-sulfate reductase (glutat	1.8.4.9
A0A078FUH8	FF153	Brassica napus	adenylyl-sulfate reductase (glutat	1.8.4.9
V4MII0	FF153	Eutrema salsugineum	adenylyl-sulfate reductase (glutat	1.8.4.9
D7ME98	FF153	Arabidopsis lyrata subsp. lyra	adenylyl-sulfate reductase (glutat	1.8.4.9
R0H0Z7	FF153	Capsella rubella	adenylyl-sulfate reductase (glutat	1.8.4.9
M4DAR9	FF153	Brassica campestris	adenylyl-sulfate reductase (glutat	1.8.4.9
M4EL38	FF153	Brassica campestris	adenylyl-sulfate reductase (glutat	1.8.4.9
W1MKF2	FF153	—	—	—
A0A0C7CX40	FF153	—	—	—
A0A1C7BU64	FF153	—	—	—
Q02KP7	FF153	Pseudomonas aeruginosa (strain	Adenosine 5'-phosphosulfate reduct	1.8.4.10
B7VBC3	FF153	Pseudomonas aeruginosa (strain	Adenosine 5'-phosphosulfate reduct	1.8.4.10
A0A335EVG0	FF153	—	—	—
V6AHA6	FF153	—	—	—
A0A072ZT75	FF153	—	—	—

Expanded dict

← click any AC on the left to see its full dict

📊 Full Excel

funfam_annotation.xlsx — per-member wide table + 4 N×N matrices × 2 FFs + decision + descriptions

download xlsx

📁 Full JSON dict

member_facts.json — one full dict per UniProt AC (all 8 sources)

download json

      In one line: input = FunFam ID; output = pure → 1 description, mixed → N. The report shows examples and evidence; the full raw data lives in the Excel.
    