{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/from-genome-to-phenome-predicting-multiple","title":"From genome to phenome: Predicting multiple cancer phenotypes based on somatic genomic alterations via the genomic impact transformer","arxiv_id":"1902.00078","date":"2019-01-31","proceeding":null,"authors":["Yifeng Tao","Chunhui Cai","William Cohen","Xinghua Lu"],"abstract":"Motivation: Cancers are mainly caused by somatic genomic alterations (SGAs)\nthat perturb cellular signaling systems and eventually activate oncogenic\nprocesses. Therefore, understanding the functional impact of SGAs is a\nfundamental task in cancer biology and precision oncology. Here, we present a\ndeep neural network model with encoder-decoder architecture, referred to as\ngenomic impact transformer (GIT), to infer the functional impact of SGAs on\ncellular signaling systems through modeling the statistical relationships\nbetween SGA events and differentially expressed genes (DEGs) in tumors. The\nmodel utilizes multi-head self-attention mechanism to identify SGAs that likely\ncause DEGs, or in other words differentiating potential driver SGAs from\npassenger ones in a tumor. GIT model learns a vector (gene embedding) as an\nabstract representation of functional impact for each SGA-affected gene. Given\nSGAs of a tumor, the model can instantiate the states of the hidden layer,\nproviding abstract representation (tumor embedding) reflecting characteristics\nof perturbed molecular/cellular processes in the tumor, which in turn can be\nused to predict multiple phenotypes. Results: We apply the GIT model to 4,468\ntumors profiled by The Cancer Genome Atlas (TCGA) project. The attention\nmechanism enables the model to better capture the statistical relationship\nbetween SGAs and DEGs than conventional methods, and distinguishes cancer\ndrivers from passengers. The learned gene embeddings capture the functional\nsimilarity of SGAs perturbing common pathways. The tumor embeddings are shown\nto be useful for tumor status representation, and phenotype prediction\nincluding patient survival time and drug response of cancer cell lines.","url_abs":"http://arxiv.org/abs/1902.00078v1","url_pdf":"http://arxiv.org/pdf/1902.00078v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"from-genome-to-phenome-predicting-multiple","repo_url":"https://github.com/yifengtao/genome-transformer","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}