IEP corpus survey
Generated by projects/IEP/iep_corpus_survey.py. Two views: raw GOA (*-goa.tsv, what GOA ships) and reviewed (*-ai-review.yaml, what reviewers concluded).
GO closure computed against GO release 2026-07-26 (OAK adapter sqlite:obo:go by default; override with IEP_GO_ADAPTER). The coarse branch tallies can move by a row or two between GO releases, as terms are obsoleted or reparented; everything else is release-independent.
The disposition table was refreshed on 2026-09-13 by recounting the checked-out gene-review YAML files. Its totals and existing action counts are unchanged; the previously omitted NEW rows are now displayed. GO-dependent sections retain the recorded snapshot above.
What the reviewer dispositions are. Every action counted below comes from this repository's own *-ai-review.yaml files, which are AI-generated reviews produced under the guidance in CLAUDE.md. They are not independent adjudication of GO curators. See the limitations note.
GOA view
- Annotation rows across all cached GOA files: 149,995
- IEP rows: 555 (0.4% of all rows), in 220 gene directories
Evidence-code frequency (top 15)
| Code |
Rows |
Share |
| IPI |
41,122 |
27.4% |
| IEA |
31,258 |
20.8% |
| IDA |
22,031 |
14.7% |
| TAS |
15,290 |
10.2% |
| IMP |
9,839 |
6.6% |
| IBA |
9,695 |
6.5% |
| ISS |
5,373 |
3.6% |
| ISO |
4,605 |
3.1% |
| NAS |
2,792 |
1.9% |
| EXP |
2,287 |
1.5% |
| HDA |
2,138 |
1.4% |
| IGI |
1,860 |
1.2% |
| IEP |
555 |
0.4% |
| HTP |
322 |
0.2% |
| IC |
252 |
0.2% |
IEP by GO aspect
| Aspect |
Rows |
Share of IEP |
| biological_process |
532 |
95.9% |
| cellular_component |
20 |
3.6% |
| molecular_function |
3 |
0.5% |
IEP by qualifier
| Qualifier |
Rows |
Share of IEP |
| involved_in |
400 |
72.1% |
| acts_upstream_of_or_within |
128 |
23.1% |
| is_active_in |
20 |
3.6% |
| enables |
3 |
0.5% |
| NOT |
involved_in |
2 |
| acts_upstream_of_negative_effect |
2 |
0.4% |
IEP by assigning group (top 15)
| Assigned by |
Rows |
Share of IEP |
| RGD |
178 |
32.1% |
| UniProt |
160 |
28.8% |
| TAIR |
95 |
17.1% |
| WB |
27 |
4.9% |
| dictyBase |
23 |
4.1% |
| SynGO |
22 |
4.0% |
| BHF-UCL |
15 |
2.7% |
| MGI |
14 |
2.5% |
| FlyBase |
7 |
1.3% |
| EcoliWiki |
4 |
0.7% |
| DisProt |
3 |
0.5% |
| TIGR |
2 |
0.4% |
| EcoCyc |
2 |
0.4% |
| HGNC-UCL |
1 |
0.2% |
| ParkinsonsUK-UCL |
1 |
0.2% |
IEP by taxon (top 15)
| Taxon |
Rows |
| Rattus norvegicus |
190 |
| Arabidopsis thaliana |
154 |
| Homo sapiens |
101 |
| Caenorhabditis elegans |
33 |
| Mus musculus |
27 |
| Dictyostelium discoideum |
23 |
| Drosophila melanogaster |
7 |
| Nicotiana attenuata |
6 |
| Escherichia coli (strain K12) |
5 |
| Oryza sativa subsp. japonica |
4 |
| Medicago truncatula |
3 |
| ? |
1 |
| Saccharomyces cerevisiae (strain ATCC 204508 / S288c) |
1 |
Most frequent IEP terms in GOA (top 25)
Review view
- Reviewed IEP rows: 550 across 221 gene review files
- IEP rows whose exact GO term is also carried by a non-IEP annotation in the same review: 152 (27.6%)
- IEP rows where IEP is the sole carrier of that GO term: 398 (72.4%), of which 82 were REMOVE/MARK_OVER/MODIFY
- Same measurement with is_a/part_of closure (an ancestor or descendant of the IEP term under a non-IEP code also counts as corroboration): 256 corroborated (46.5%), so the sole-carrier share falls to 53.5%. The exact-match figure above is therefore an upper bound on sole carriage.
- IEP rows whose term also appears in the review's
core_functions: 55 (10.0%), of which 50 were also ACCEPTed (the rest reach core_functions on the strength of a co-annotated non-IEP row)
How concentrated IEP is over genes
- Gene directories carrying at least one reviewed IEP row: 221; median IEP rows per such gene: 1; maximum: 33
- The top 10% of IEP-carrying genes (23 genes) carry 235 rows, 42.7% of the total
- Genes with 5 or more IEP rows: 21 (9.5% of IEP-carrying genes), carrying 41.3% of all IEP rows (this is the statistic comparable to the global atlas)
Disposition by evidence code
| Code |
Reviewed |
ACCEPT |
KEEP_AS_NON_CORE |
MODIFY |
MARK_AS_OVER_ANNOTATED |
REMOVE |
UNDECIDED |
PENDING |
NEW |
UNREVIEWED |
% negative |
| IDA |
21,893 |
15167 |
4290 |
612 |
635 |
233 |
244 |
47 |
665 |
0 |
6.8% |
| IMP |
9,863 |
4766 |
3944 |
252 |
327 |
96 |
121 |
10 |
347 |
0 |
6.8% |
| IGI |
1,557 |
706 |
699 |
39 |
48 |
12 |
36 |
2 |
15 |
0 |
6.4% |
| IPI |
17,834 |
1916 |
3395 |
1493 |
7836 |
2994 |
106 |
15 |
79 |
0 |
69.1% |
| IEP |
550 |
124 |
306 |
7 |
74 |
11 |
3 |
22 |
3 |
0 |
16.7% |
| IBA |
9,728 |
7037 |
1521 |
422 |
390 |
263 |
42 |
11 |
42 |
0 |
11.1% |
| ISO |
4,245 |
1406 |
1980 |
114 |
381 |
218 |
128 |
0 |
18 |
0 |
16.8% |
| IEA |
31,612 |
15998 |
8331 |
2047 |
3318 |
1072 |
282 |
250 |
314 |
0 |
20.4% |
| TAS |
15,432 |
11408 |
2906 |
261 |
514 |
161 |
40 |
0 |
142 |
0 |
6.1% |
| NAS |
3,021 |
1706 |
607 |
141 |
263 |
48 |
14 |
0 |
242 |
0 |
15.0% |
% negative = REMOVE + MARK_AS_OVER_ANNOTATED + MODIFY, i.e. rows a reviewer judged not keepable as written, over the Reviewed total. That total is every annotation row carrying the code, including reviewer-proposed NEW annotations and PENDING/UNREVIEWED rows, so the action columns sum to it. NEW rows are proposals from this project, not dispositions of an existing GOA annotation.
How often each code lands on core function
| Code |
Reviewed |
ACCEPT |
% ACCEPT |
Term in core_functions |
% core |
| IDA |
21,893 |
15,167 |
69.3% |
9,616 |
43.9% |
| IMP |
9,863 |
4,766 |
48.3% |
3,237 |
32.8% |
| IGI |
1,557 |
706 |
45.3% |
414 |
26.6% |
| IPI |
17,834 |
1,916 |
10.7% |
768 |
4.3% |
| IEP |
550 |
124 |
22.5% |
55 |
10.0% |
| IBA |
9,728 |
7,037 |
72.3% |
4,944 |
50.8% |
| ISO |
4,245 |
1,406 |
33.1% |
687 |
16.2% |
| IEA |
31,612 |
15,998 |
50.6% |
9,299 |
29.4% |
| TAS |
15,432 |
11,408 |
73.9% |
8,666 |
56.2% |
| NAS |
3,021 |
1,706 |
56.5% |
1,305 |
43.2% |
% core credits a code whenever the term it carries also appears in core_functions, even if the term got there on the strength of a different code annotating it too. It is therefore generous to every code, and most generous to codes that frequently co-annotate.
What IEP is used to say (GO branch, is_a + part_of closure)
| Branch |
IEP rows |
Share |
Genes |
Flagged |
% flagged |
Genes flagged |
| response to stimulus |
381 |
69.3% |
153 |
64 |
16.8% |
32 |
| developmental process |
84 |
15.3% |
54 |
19 |
22.6% |
17 |
| biological regulation |
33 |
6.0% |
27 |
5 |
15.2% |
5 |
| unclassified |
24 |
4.4% |
21 |
2 |
8.3% |
2 |
| cellular component |
14 |
2.5% |
8 |
0 |
0.0% |
0 |
| localization |
6 |
1.1% |
3 |
0 |
0.0% |
0 |
| metabolic process |
5 |
0.9% |
5 |
1 |
20.0% |
1 |
| molecular function |
3 |
0.5% |
3 |
1 |
33.3% |
1 |
Branch assignment is first match wins in the order listed in BRANCHES, with response to stimulus tested before developmental process, so a term parented under both is counted as stimulus-response.
Branch flag rates, split by species
The pooled branch comparison above can be confounded by review batch: if one species dominates one branch, the branch flag rate may be measuring that species' review batch instead. This table splits the two largest branches by species and gives a two-sided Fisher exact test of the developmental-versus-stimulus flag-rate difference within each.
| Species |
response to stimulus |
flagged |
% |
developmental process |
flagged |
% |
Fisher p |
| rat |
136 |
34 |
25.0% |
27 |
9 |
33.3% |
0.473 |
| ARATH |
128 |
11 |
8.6% |
5 |
2 |
40.0% |
0.075 |
| human |
48 |
5 |
10.4% |
34 |
7 |
20.6% |
0.221 |
| worm |
27 |
0 |
0.0% |
0 |
0 |
n/a |
1.000 |
| mouse |
8 |
8 |
100.0% |
4 |
0 |
0.0% |
0.002 |
| DICDI |
11 |
2 |
18.2% |
10 |
1 |
10.0% |
1.000 |
| DROME |
6 |
0 |
0.0% |
1 |
0 |
0.0% |
1.000 |
| NICAT |
6 |
0 |
0.0% |
0 |
0 |
n/a |
1.000 |
| ECOLI |
5 |
1 |
20.0% |
0 |
0 |
n/a |
1.000 |
| ORYSJ |
4 |
3 |
75.0% |
0 |
0 |
n/a |
1.000 |
| MEDTR |
0 |
0 |
n/a |
3 |
0 |
0.0% |
1.000 |
| TOBAC |
2 |
0 |
0.0% |
0 |
0 |
n/a |
1.000 |
| all |
381 |
64 |
16.8% |
84 |
19 |
22.6% |
0.210 |
Pooled, the developmental branch flags at 22.6% against 16.8% for stimulus-response, on 19/84 versus 64/381 (two-sided Fisher p = 0.21). The developmental rows span 54 gene directories. These are annotation-level comparisons: rows can share genes, references, and review batches. The p-values are unadjusted, and neither a pooled contrast nor the number of genes rules out those dependencies or establishes a branch-wide difference in annotation quality.
Reviewed IEP rows by species (top 15)
| Species |
IEP rows |
| rat |
185 |
| ARATH |
154 |
| human |
101 |
| worm |
34 |
| mouse |
25 |
| DICDI |
23 |
| DROME |
7 |
| NICAT |
6 |
| ECOLI |
5 |
| ORYSJ |
4 |
| MEDTR |
3 |
| TOBAC |
2 |
| yeast |
1 |
Most frequently flagged IEP terms
All flagged IEP rows
| Species |
Gene |
GO term |
Label |
Action |
| ARATH |
AG |
GO:0048366 |
leaf development |
REMOVE |
| ARATH |
AT5G03720 |
GO:0009408 |
response to heat |
MODIFY |
| ARATH |
CRY1 |
GO:0009416 |
response to light stimulus |
MODIFY |
| ARATH |
CRY2 |
GO:0009416 |
response to light stimulus |
MODIFY |
| ARATH |
PAD4 |
GO:0051707 |
response to other organism |
MODIFY |
| ARATH |
PIF3 |
GO:0009408 |
response to heat |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:0009409 |
response to cold |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:0009723 |
response to ethylene |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:0009733 |
response to auxin |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:0009737 |
response to abscisic acid |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:1902074 |
response to salt |
MARK_AS_OVER_ANNOTATED |
| ARATH |
PIF3 |
GO:1990785 |
response to water-immersion restraint stress |
MARK_AS_OVER_ANNOTATED |
| ARATH |
SOC1 |
GO:0045893 |
positive regulation of DNA-templated transcription |
MODIFY |
| ARATH |
WRKY70 |
GO:0010150 |
leaf senescence |
MODIFY |
| DICDI |
acaA |
GO:1902351 |
response to imidacloprid |
MARK_AS_OVER_ANNOTATED |
| DICDI |
cotB |
GO:0031153 |
slug development involved in sorocarp development |
MARK_AS_OVER_ANNOTATED |
| DICDI |
mhcA |
GO:0031152 |
aggregation involved in sorocarp development |
MARK_AS_OVER_ANNOTATED |
| ECOLI |
arnF |
GO:0010041 |
response to iron(III) ion |
MARK_AS_OVER_ANNOTATED |
| ORYSJ |
EME1 |
GO:0009411 |
response to UV |
REMOVE |
| ORYSJ |
EME1 |
GO:0009644 |
response to high light intensity |
REMOVE |
| ORYSJ |
EME1 |
GO:0010332 |
response to gamma radiation |
REMOVE |
| human |
ACADVL |
GO:0030855 |
epithelial cell differentiation |
MARK_AS_OVER_ANNOTATED |
| human |
ACTA2 |
GO:0009615 |
response to virus |
MARK_AS_OVER_ANNOTATED |
| human |
ACTA2 |
GO:0072144 |
glomerular mesangial cell development |
MARK_AS_OVER_ANNOTATED |
| human |
ACTL8 |
GO:0030855 |
epithelial cell differentiation |
MARK_AS_OVER_ANNOTATED |
| human |
BAG6 |
GO:0140677 |
molecular function activator activity |
REMOVE |
| human |
CDK1 |
GO:0030855 |
epithelial cell differentiation |
MARK_AS_OVER_ANNOTATED |
| human |
CLU |
GO:0009615 |
response to virus |
MARK_AS_OVER_ANNOTATED |
| human |
CPT1A |
GO:0030855 |
epithelial cell differentiation |
MARK_AS_OVER_ANNOTATED |
| human |
FN3K |
GO:0030855 |
epithelial cell differentiation |
REMOVE |
| human |
IFI16 |
GO:0010506 |
regulation of autophagy |
MARK_AS_OVER_ANNOTATED |
| human |
MYC |
GO:0009410 |
response to xenobiotic stimulus |
MARK_AS_OVER_ANNOTATED |
| human |
MYC |
GO:0034644 |
cellular response to UV |
MARK_AS_OVER_ANNOTATED |
| human |
PGK1 |
GO:0030855 |
epithelial cell differentiation |
MARK_AS_OVER_ANNOTATED |
| human |
RB1 |
GO:0007265 |
Ras protein signal transduction |
REMOVE |
| mouse |
Ifi204 |
GO:0009617 |
response to bacterium |
MODIFY |
| mouse |
Mir100 |
GO:0007605 |
sensory perception of sound |
REMOVE |
| mouse |
Mir100 |
GO:0060291 |
long-term synaptic potentiation |
REMOVE |
| mouse |
Mir100 |
GO:1990830 |
cellular response to leukemia inhibitory factor |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir127 |
GO:0060291 |
long-term synaptic potentiation |
REMOVE |
| mouse |
Mir127 |
GO:1990830 |
cellular response to leukemia inhibitory factor |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir26a-1 |
GO:0009611 |
response to wounding |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir26a-1 |
GO:0009617 |
response to bacterium |
REMOVE |
| mouse |
Mir30e |
GO:0009617 |
response to bacterium |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir30e |
GO:0060291 |
long-term synaptic potentiation |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir30e |
GO:0071361 |
cellular response to ethanol |
MARK_AS_OVER_ANNOTATED |
| mouse |
Mir384 |
GO:1990830 |
cellular response to leukemia inhibitory factor |
MARK_AS_OVER_ANNOTATED |
| rat |
Akt1 |
GO:1901653 |
cellular response to peptide |
MARK_AS_OVER_ANNOTATED |
| rat |
Aprt |
GO:0032869 |
cellular response to insulin stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Casp3 |
GO:0036269 |
swimming behavior |
MARK_AS_OVER_ANNOTATED |
| rat |
Ckmt2 |
GO:0007507 |
heart development |
MARK_AS_OVER_ANNOTATED |
| rat |
Ckmt2 |
GO:0007519 |
skeletal muscle tissue development |
MARK_AS_OVER_ANNOTATED |
| rat |
Ephx1 |
GO:0001889 |
liver development |
MARK_AS_OVER_ANNOTATED |
| rat |
Ephx1 |
GO:0071385 |
cellular response to glucocorticoid stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Gamt |
GO:1990402 |
embryonic liver development |
MARK_AS_OVER_ANNOTATED |
| rat |
Ghr |
GO:0009629 |
response to gravity |
MARK_AS_OVER_ANNOTATED |
| rat |
Gss |
GO:0009410 |
response to xenobiotic stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Gss |
GO:0031667 |
response to nutrient levels |
MARK_AS_OVER_ANNOTATED |
| rat |
Gss |
GO:0034612 |
response to tumor necrosis factor |
MARK_AS_OVER_ANNOTATED |
| rat |
Gss |
GO:0043200 |
response to amino acid |
MARK_AS_OVER_ANNOTATED |
| rat |
Gsta4 |
GO:0009635 |
response to herbicide |
MARK_AS_OVER_ANNOTATED |
| rat |
Gsta4 |
GO:0010043 |
response to zinc ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Gsta4 |
GO:0035094 |
response to nicotine |
MARK_AS_OVER_ANNOTATED |
| rat |
Gsta4 |
GO:0071285 |
cellular response to lithium ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Gstt1 |
GO:0009410 |
response to xenobiotic stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Gstt1 |
GO:0009751 |
response to salicylic acid |
MARK_AS_OVER_ANNOTATED |
| rat |
Gstt1 |
GO:0010269 |
response to selenium ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Gstt1 |
GO:0033197 |
response to vitamin E |
MARK_AS_OVER_ANNOTATED |
| rat |
Hmgcs2 |
GO:0009266 |
response to temperature stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Hmgcs2 |
GO:0030324 |
lung development |
MARK_AS_OVER_ANNOTATED |
| rat |
Hmgcs2 |
GO:0060612 |
adipose tissue development |
MARK_AS_OVER_ANNOTATED |
| rat |
Hmgcs2 |
GO:0071230 |
cellular response to amino acid stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0001666 |
response to hypoxia |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0009410 |
response to xenobiotic stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0032094 |
response to food |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0032868 |
response to insulin |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0048545 |
response to steroid hormone |
MARK_AS_OVER_ANNOTATED |
| rat |
Hsd11b2 |
GO:0051384 |
response to glucocorticoid |
MARK_AS_OVER_ANNOTATED |
| rat |
Oaz1 |
GO:0002931 |
response to ischemia |
MARK_AS_OVER_ANNOTATED |
| rat |
Oaz1 |
GO:0090650 |
cellular response to oxygen-glucose deprivation |
MARK_AS_OVER_ANNOTATED |
| rat |
Pgam2 |
GO:0007283 |
spermatogenesis |
MARK_AS_OVER_ANNOTATED |
| rat |
Pnlip |
GO:0009791 |
post-embryonic development |
MARK_AS_OVER_ANNOTATED |
| rat |
Pnlip |
GO:0033993 |
response to lipid |
MARK_AS_OVER_ANNOTATED |
| rat |
Pnlip |
GO:0043434 |
response to peptide hormone |
MARK_AS_OVER_ANNOTATED |
| rat |
Qdpr |
GO:0001889 |
liver development |
MARK_AS_OVER_ANNOTATED |
| rat |
Qdpr |
GO:0010044 |
response to aluminum ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Qdpr |
GO:0010288 |
response to lead ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Qdpr |
GO:0033762 |
response to glucagon |
MARK_AS_OVER_ANNOTATED |
| rat |
Qdpr |
GO:0071466 |
cellular response to xenobiotic stimulus |
MARK_AS_OVER_ANNOTATED |
| rat |
Ube2d2b |
GO:0071276 |
cellular response to cadmium ion |
MARK_AS_OVER_ANNOTATED |
| rat |
Ube2d2b |
GO:1903841 |
cellular response to arsenite(3-) |
MARK_AS_OVER_ANNOTATED |
| yeast |
THI22 |
GO:0009228 |
thiamine biosynthetic process |
MARK_AS_OVER_ANNOTATED |