Fig. 3: The protein–protein interaction network of the genes containing the AI-assisted identified SNPs. | npj Digital Medicine

Fig. 3: The protein–protein interaction network of the genes containing the AI-assisted identified SNPs.

From: Discovering a trans-omics biomarker signature that predisposes high risk diabetic patients to diabetic kidney disease

Fig. 3

The protein-coding genes containing the SNPs identified by AI were used to build up a protein–protein interaction network. Ranking of the protein-coding genes in the feature importance of the three models: Model 1, RAPTOR (rank 23) and CLPTM1L (rank 25); Model 2, PCDH9 (rank 9) and ALDH1L1 (rank 7); Model 3, FAM53A (rank 19), ADCYAP (rank 16), B3GNTL1 (rank 10) and CDS1 (rank 13). The figure was created using the open-source software Cytoscape. CLPTM1L cleft lip and palate transmembrane protein 1-like, RPTOR regulatory associated protein of mTOR complex 1, ALDH1L1 aldehyde dehydrogenase 1 family member L1, PCDH9 protocadherin 9, B3GNTL1 UDP-GlcNAc:BetaGal Beta-1,3-N-Acetylglucosaminyltransferase like 1, CDS1 CDP-Diacylglycerol synthase 1, FAM53A family with sequence similarity 53 member A, ADCYAP1 adenylate cyclase activating polypeptide 1, RNF4 ring finger protein 4, USP9X, ubiquitin specific peptidase 9 X-linked, HSP90AA1 heat shock protein 90 alpha family class a member 1, PSMA4 proteasome 20 S subunit alpha 4, HNRNPL heterogeneous nuclear ribonucleoprotein L.

Back to article page