Study
Enformer Recognizes cis-Regulated Genes by Prediction Magnitude but Fails to Predict Effect Direction: A Magnitude–Direction Dissociation in Individual-Level Expression Prediction
View full titleShow less
We cross-reference Enformer's individual-level per-gene Pearson R (Sasse et al.
Abstract
We cross-reference Enformer's individual-level per-gene Pearson R (Sasse et al. 2023, Nature Genetics, ROSMAP cohort n=839) with GTEx v8 Brain Cortex eGenes (n=9,082) across 6,808 genes, and report a fundamental dissociation: prediction magnitude (|R|) monotonically predicts cis-eQTL gene identity (eGene rate 31.6% → 82.5%; |R| model AIC=8274 vs R+R² AIC=8378, ΔAIC=104), but prediction direction for eGenes is near-random (54.0% correct; max D10=62.5%). Direction errors are irreversible by fine-tuning on ROSMAP (Mann–Whitney p≈10⁻¹⁰⁰), yet PrediXcan recovers correct direction for R<0 eGenes (PrediXcanR=0.295), consistent with the hypothesis that Enformer's non-brain ENCODE training corpus lacks the brain-specific regulatory programs needed to correctly orient directional effects.
This page presents preserved, source-backed released materials.
Source inventory
Deterministically extracted immutable study materials.
Extracted study record
Authoritative metadata and released result artifacts.
The preserved source contains 5 result tables; 5 have complete deterministic representation coverage.
Gene Ids And Positions source data
58302 materialized released rows from Data/gene-ids-and-positions.tsv.
Evidence visualization
Gene Ids And Positions source data
58302 materialized released rows from Data/gene-ids-and-positions.tsv.
Browse source data · 58,302 rows
Rows 1–10 of 58302
| Gene Name | Gene Id | Chr Hg38 | Start Hg38 | End Hg38 | Strand Hg38 | Tss Hg38 | Chr Hg19 | Tss Hg19 |
|---|---|---|---|---|---|---|---|---|
| DDX11L1 | ENSG00000223972 | 1 | 11869 | 14409 | + | 11869 | chr1 | 11869 |
| WASH7P | ENSG00000227232 | 1 | 14404 | 29570 | - | 29570 | chr1 | 29570 |
| MIR6859-1 | ENSG00000278267 | 1 | 17369 | 17436 | - | 17436 | chr1 | 17436 |
| MIR1302-2HG | ENSG00000243485 | 1 | 29554 | 31109 | + | 29554 | chr1 | 29554 |
| MIR1302-2 | ENSG00000284332 | 1 | 30366 | 30503 | + | 30366 | chr1 | 30366 |
| FAM138A | ENSG00000237613 | 1 | 34554 | 36081 | - | 36081 | chr1 | 36081 |
| OR4G4P | ENSG00000268020 | 1 | 52473 | 53312 | + | 52473 | chr1 | 52473 |
| OR4G11P | ENSG00000240361 | 1 | 57598 | 64116 | + | 57598 | chr1 | 57598 |
| OR4F5 | ENSG00000186092 | 1 | 65419 | 71585 | + | 65419 | chr1 | 65419 |
| AL627309.1 | ENSG00000238009 | 1 | 89295 | 133723 | - | 133723 | chr1 | 133723 |
Source & verification
- File
- Data/gene-ids-and-positions.tsv
- Version
afac6658- Integrity
1c1bf881…a9a24f
PrediXcanCorrelationWithExpressionENSG source data
6824 materialized released rows from Data/PrediXcanCorrelationWithExpressionENSG.tsv.
Evidence visualization
PrediXcanCorrelationWithExpressionENSG source data
6824 materialized released rows from Data/PrediXcanCorrelationWithExpressionENSG.tsv.
Browse source data · 6,824 rows
Rows 1–10 of 6824
| ENSG00000000419 | NaN |
|---|---|
| ENSG00000000457 | 0.205679 |
| ENSG00000000938 | NaN |
| ENSG00000000971 | NaN |
| ENSG00000001036 | NaN |
| ENSG00000001084 | 0.327986 |
| ENSG00000001167 | NaN |
| ENSG00000001460 | 0.287331 |
| ENSG00000001461 | NaN |
| ENSG00000001561 | 0.365575 |
| ENSG00000001617 | NaN |
Source & verification
- File
- Data/PrediXcanCorrelationWithExpressionENSG.tsv
- Version
afac6658- Integrity
8681bac2…889ad1
SupplementaryTable1 source data
6820 materialized released rows from Data/SupplementaryTable1.tsv.
Evidence visualization
SupplementaryTable1 source data
6820 materialized released rows from Data/SupplementaryTable1.tsv.
Browse source data · 6,820 rows
Rows 1–10 of 6820
| # Gene Name | Gene Id | Chr Hg38 | Start Hg38 | End Hg38 | Strand Hg38 | Tss Hg38 | MeanObs | StdObs | PearsonRfineTuned | PvalueRfineTuned | MeanFineTuned | StdFineTuned | PearsonRCAGE,Adult,Brain | PvalueRCAGE,Adult,Brain | MeanCAGE,Adult,Brain | StdCAGE,Adult,Brain | RmeanRandom | Rstdrandom | TstatCAGE,Adult,Brain |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DPM1 | ENSG00000000419 | 20 | 50934867 | 50958555 | - | 50958555 | 3.43 | 0.25 | 0.031 | 0.373 | 8.49 | 0.026 | 0.068 | 0.047 | 2.21 | 0.013 | 0.028 | 0.02213 | 1.901 |
| SCYL3 | ENSG00000000457 | 1 | 169849631 | 169894267 | - | 169894267 | 3.02 | 0.264 | -0.028 | 0.42 | 5.73 | 0.014 | 0.177 | 0.0 | 0.28 | 0.005 | 0.132 | 0.0706 | 1.198 |
| FGR | ENSG00000000938 | 1 | 27612064 | 27635277 | - | 27635277 | 2.48 | 0.528 | 0.039 | 0.26 | 5.15 | 0.048 | 0.091 | 0.008 | 0.78 | 0.012 | 0.022 | 0.01594 | 3.864 |
| CFH | ENSG00000000971 | 1 | 196651878 | 196747504 | + | 196651878 | 4.18 | 0.725 | 0.089 | 0.01 | 4.24 | 0.036 | 0.037 | 0.279 | 1.43 | 0.01 | 0.048 | 0.03346 | -0.425 |
| FUCA2 | ENSG00000001036 | 6 | 143494811 | 143511690 | - | 143511690 | 3.05 | 0.303 | -0.031 | 0.374 | 7.02 | 0.028 | -0.022 | 0.524 | 2.57 | 0.01 | 0.029 | 0.02018 | -0.348 |
| GCLC | ENSG00000001084 | 6 | 53497341 | 53616970 | - | 53616970 | 4.69 | 0.275 | -0.081 | 0.018 | 1.96 | 0.041 | -0.051 | 0.143 | 0.1 | 0.014 | 0.054 | 0.03873 | -0.108 |
| NFYA | ENSG00000001167 | 6 | 41072945 | 41099976 | + | 41072945 | 4.07 | 0.218 | 0.085 | 0.014 | 6.53 | 0.012 | 0.112 | 0.001 | 0.83 | 0.007 | 0.058 | 0.03596 | 2.014 |
| STPG1 | ENSG00000001460 | 1 | 24356999 | 24416934 | - | 24416934 | 3.89 | 0.248 | 0.093 | 0.007 | 4.86 | 0.042 | -0.226 | 0.0 | 0.05 | 0.002 | 0.084 | 0.0573 | 4.195 |
| NIPAL3 | ENSG00000001461 | 1 | 24415794 | 24472976 | + | 24415794 | 6.45 | 0.208 | -0.027 | 0.428 | 9.51 | 0.065 | -0.015 | 0.667 | 2.48 | 0.019 | 0.02 | 0.0124 | -0.318 |
| ENPP4 | ENSG00000001561 | 6 | 46129993 | 46146699 | + | 46129993 | 4.92 | 0.301 | -0.191 | 0.0 | 8.95 | 0.173 | -0.2 | 0.0 | 4.08 | 0.089 | 0.106 | 0.05915 | 2.733 |
Source & verification
- File
- Data/SupplementaryTable1.tsv
- Version
afac6658- Integrity
86a0dbe9…cd4009
SupplementaryTable2 source data
1091 materialized released rows from Data/SupplementaryTable2.tsv.
Evidence visualization
SupplementaryTable2 source data
1091 materialized released rows from Data/SupplementaryTable2.tsv.
Browse source data · 1,091 rows
Rows 1–10 of 1091
| # Gene Name | Gene Id | Chr Hg38 | Start Hg38 | End Hg38 | Strand Hg38 | Tss Hg38 | LociDriverhg38 | DisttoTSS | ISM | SumISMPearsonRtoCAGE,Adult,Brain | DriverPearsonRtoCAGE,Adult,Brain | DecreasedFractionofSumPearsonRWithoutDrivertoCAGE,Adult,Brain | IncreasedFractionofSumPearonRbyrankedDrivertoCAGE,Adult,Brain | NIndividualsWithSNV | FractionIndividualsWithSNV | IsMainDriver | ISMFractionofISMmaxinSNV | EQTL | MaxeQTLgene |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| STPG1 | ENSG00000001460 | 1 | 24356999 | 24416934 | - | 24416934 | 24417389 | 455 | 0.01265 | 0.97 | 0.97 | 1.25 | 1.0 | 509 | 43.8 | True | 1.0 | -0.107638 | 1.05414 |
| FARP2 | ENSG00000006607 | 2 | 241356243 | 241494841 | + | 241356243 | 241356233 | -10 | -0.99268 | 0.92 | 0.94 | 0.83 | 1.02 | 1114 | 96.0 | True | -0.202 | 0.128377 | 0.870283 |
| DDX11 | ENSG00000013573 | 12 | 31073845 | 31104791 | + | 31073845 | 31073901 | 56 | -3.13881 | 0.96 | 0.95 | 2.0 | 0.99 | 799 | 68.8 | True | -1.0 | -0.923821 | 2.189899 |
| POLA2 | ENSG00000014138 | 11 | 65261762 | 65305589 | + | 65261762 | 65261966 | 204 | 1.04261 | 0.98 | 0.65 | 0.13 | 0.67 | 172 | 14.8 | True | 1.0 | 0.155294 | 1.319672 |
| POLA2 | ENSG00000014138 | 11 | 65261762 | 65305589 | + | 65261762 | 65266403 | 4641 | -0.55469 | 0.98 | 0.96 | 0.41 | 0.32 | 313 | 27.0 | False | -0.532 | -0.065785 | 1.319672 |
| STMN4 | ENSG00000015592 | 8 | 27235323 | 27258420 | - | 27258420 | 27245507 | -12913 | -3.74149 | 0.9 | 0.82 | 0.32 | 0.39 | 703 | 60.6 | False | -0.862 | -0.192287 | 1.937384 |
| STMN4 | ENSG00000015592 | 8 | 27235323 | 27258420 | - | 27258420 | 27258225 | -195 | 2.83392 | 0.9 | 0.87 | 0.05 | 0.05 | 114 | 9.8 | False | 0.653 | 0.050181 | 1.937384 |
| STMN4 | ENSG00000015592 | 8 | 27235323 | 27258420 | - | 27258420 | 27261676 | 3256 | -4.33811 | 0.9 | 0.46 | 0.01 | 0.52 | 68 | 5.9 | True | -1.0 | -0.081375 | 1.937384 |
| ZDHHC6 | ENSG00000023041 | 10 | 112424428 | 112446917 | - | 112446917 | 112430594 | -16323 | -0.11328 | 0.82 | 0.71 | 0.06 | 0.07 | 248 | 21.4 | False | -0.072 | -0.11775 | 0.763484 |
| ZDHHC6 | ENSG00000023041 | 10 | 112424428 | 112446917 | - | 112446917 | 112446996 | 79 | -0.47361 | 0.82 | 0.83 | 0.14 | 0.83 | 1064 | 91.6 | True | -0.3 | -0.026299 | 0.763484 |
Source & verification
- File
- Data/SupplementaryTable2.tsv
- Version
afac6658- Integrity
1618b841…3fa458
GTExWeights source data
5314 materialized released rows from fine_tuned_model/GTExWeights.csv.
Evidence visualization
GTExWeights source data
5314 materialized released rows from fine_tuned_model/GTExWeights.csv.
Browse source data · 5,314 rows
Rows 1–10 of 5314
| Tracks | Weights |
|---|---|
| DNASE:cerebellum male adult (27 years) and male adult (35 years) | 0 |
| DNASE:frontal cortex male adult (27 years) and male adult (35 years) | 0 |
| DNASE:chorion | 0 |
| DNASE:Ishikawa treated with 0.02% dimethyl sulfoxide for 1 hour | -0.021459 |
| DNASE:GM03348 | 0 |
| DNASE:GM03348 genetically modified using transduction treated with 3 ug/mL doxycycline for 10 days | 0 |
| DNASE:AG08395 | 0 |
| DNASE:AG08396 | 0 |
| DNASE:AG20443 | 0 |
| DNASE:frontal cortex female adult (67 years) and female adult (80 years) | 0 |
Source & verification
- File
- fine_tuned_model/GTExWeights.csv
- Version
afac6658- Integrity
68e68810…ebd21c
Deterministic extraction method
No generative model selected, summarized, or transformed the results.
Evidence classified preserved repository files, parsed bounded structured tables, inferred column types, and selected registered components using versioned rules.
Data and provenance
Immutable publication and repository sources.
Reuse boundary
Only released and preserved source rows are displayed.
No model is rerun and no missing value is inferred.
Pinned scientific commit: afac6658d57bfe6e2c1f17d9a96e3593843000a2
Limitations
Deterministic coverage boundaries.
Every accepted released table cell is represented by registered Evidence components; unstructured claims are not converted into numerical results.
Citation
Cite the scholarly record and pinned analysis source.
Scholarly publication
DongKoo Lee. Enformer Recognizes cis-Regulated Genes by Prediction Magnitude but Fails to Predict Effect Direction: A Magnitude–Direction Dissociation in Individual-Level Expression Prediction. 10.5281/zenodo.20754855.
Pinned analysis source
https://github.com/mostafavilabuw/EnformerAssessment, commit afac6658d57bfe6e2c1f17d9a96e3593843000a2.