Update: original input files by the authors are here.
Input files for the example notebook describing the finetuning process for transcription factor (TF) dosage sensitivity prediction.
| Input file | Description |
|---|---|
| Dosage sensitivity TF labels | Dosage sensitivity predictions curated by Ni et al (2019) |
| Gene info table | Ensemble gene IDs retrieved from Ensembl Genes 109 via BioMart |
Ni et al (2019) gathered dosage sensitivity prediction scores from two sources:
- High confident haploinsufficient genes (pLI>0.9) from Lek et al (2016)
- Likely dosage sensitive genes (HIPred>0.5) from Shihab et al (2017)
They also retrieved a set of all human transcription factors from Lampert et al (2018) (human TF data v1.01), and generated most reliable dosage-sensitive (MRDS) and dosage-insensitive (MRDIS) gene sets by considering support by each source dataset. Final list inclues 122 TFs with positive label (dosage sensitive, MRDS) and 368 TFs with negative label (dosage insensitive, MRDIS) out of a total of 1639 human TFs.
- Lambert, A. et al (2018) The human transcription factors. Cell 172 (4)
- Lek, M. et al (2016) Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285–291
- Shihab, A. et al (2017) HIPred: an integrative approach to predicting haploinsufficient genes. Bioinformatics 33, 1751–1757
- Ni, Z. et al (2019) Characterization of human dosage-sensitive transcription factor genes. Front. Genet. 10, 1208