Kaplan Meier survival analysis To validate our results, we collected mRNA expression profiles and clinical annotation data of patients from the TCGA website. Here, we used the mRNA expres sion profiles of three cancer types lung adenocarcinoma, colon adenocarcinoma, and skin cutaneous melanoma. The RSEM values of mRNA were used as the gene expression level measure. All P values were performed using a log rank test. Notably, for the patients of lung and colon adenocarcinoma, 2,000 day sur vival rates were used. Results Overview of somatic mutations in protein pocket regions We mapped 1,195,223 cancer related somatic mutations onto a set of 5,371 single chain proteins with pocket re gion annotations in the PDB format. The SIFTS project provided mapping information for the genomic coordinates of somatic mutations and the sequence coordinates of PDB pockets.

The final list was comprised of 2,262 unique som atic mutations in the pocket regions of 369 unique human proteins. We first examined the protein pocket region mutations at the sequence level. Among the 2,262 somatic mutations in the pocket regions, 1,603 were missense muta tions, followed by 467 silent mutations. Only a small portion of these mutations were nonsense mutations, which likely truncate protein se quences. The top 10 frequently mutated genes measured by missense mutations in the pocket regions were PIK3CA, HRAS, CRP AKT1, NCF1, NCAM2, VWF, ETV6, IFNB1, and KDM5C. It is worth noting that five of these genes are known to play important roles in cancer and are CGC genes. The average number of mutations in a pocket region per protein is 6.

1 with 4. 3 missense mutations on average per protein. For cancer types, somatic mutations in the pocket re gions were more frequently observed in uterine, skin, colon, stomach, breast, lung adenocarcinoma, head and neck, lung squamous cell, and bladder cancer than in other types. Hotspot amino acids measured by missense mutations in pocket regions We provided a catalog of amino acids involved in known somatic mutations within the pocket regions of each cancer type. This resource allows us to explore the fea tures of somatic AV-951 mutations, such as hotspot mutated amino acids in the pocket regions and their underlying mutational processes. We examined the hotspot amino acids altered by somatic mutations across 21 cancer types using COSMIC and TCGA data.

Figure 2A shows the spectrum of amino acid changes. We found that argin ine is a hotspot amino acid with a high frequency of somatic mutations in pocket regions across multiple can cer types, including uterine, skin melanoma, colon, stom ach, head and neck, and lung cancers. For example, Arg is attributed to the APOBEC family of cytidine deaminases. APOBEC3G is a member of the polynucleotide cytosine deaminase gene family, which plays important roles in anti viral immunity and cell cycles.