Insights in Computational Genomics: 2022

Recent research has identified a variety of methods to improve the diagnosis and treatment of colorectal cancer, including the identification of nine prognosis-related immune genes and a risk score to accurately predict patients' prognosis. Horizontal gene transfer mediated by conjugation has been found to be a critical force for antimicrobial resistance spread across genera. Machine learning (ML) is being used to enhance disease modeling and diagnosis through deep metabolomic profiling, while AnnotaPipeline is a computational workflow that integrates distinct software and data types to annotate and validate predicted features in genomic sequences.

The Human Pangenome Project is an effort to create a more accurate and global representation of genomic variations, but there are ethical, legal and social implications to consider. Gene duplication can be difficult to detect, but metrics and computational approaches have been developed to identify gene duplicates, as well as a BLAST-based web tool and database. Single cell RNA sequencing (scRNA-seq) is a powerful technology in biomedical research that allows for the profiling of the transcriptome of individual cells and the detection and quantitative analysis of cellular content.

Drug repositioning has been used to identify existing drugs that can be used to treat esophageal cancer patients, while deep learning (DL) and non-deep machine learning (ML) methods have been compared for predicting the risk of three lung diseases using the UK Biobank data. Nucleotide substitutions in protein-coding genes have been studied to understand the effects of purifying selection, and key seed development genes related to phytohormone biosynthesis and starch biosynthesis have been identified to improve tartary buckwheat yield. Cookie is a toolkit that can efficiently select out the most representative samples from a massive single-cell population with diverse properties.