Skip to main content

Table 2 SRA tasks

From: Standard machine learning approaches outperform deep representation learning on phenotype prediction from transcriptomics data

SRA tasks    
ProjectDiseaseLabelLabel typeGroupSamples
GSE65832atopic dermatitislesional vs. notbinarytrain40
GSE66207Crohn’s diseasetype: B1, B2 or B3multiclass (3)train20
GSE72819ulcerative colitistreatment remissionbinaryvalidate69
GSE47944psoriasislesional vs. notmulticlass (3)validate63
GSE50244diabetesnormoglycemic, impaired, diabeticmulticlass (3)validate76
GSE67785psoriasislesional vs. notbinarytest28
  1. The 8 tasks derived from SRA used to train supervised models and validate the unsupervised embeddings. The project names correspond to those in Fig. 2