Ramana Davuluri is a world leader in Molecular Data Science, with research focus on computational analysis of non-coding genomic regions and isoform-level gene regulation. Davuluri is well-known for his pioneering efforts to use machine learning in biomedical applications – for example, development of gene promoter prediction algorithms, molecular subtyping assays for glioblastoma and ovarian cancers, deep learning algorithm for integrative earning from multi-omics datasets, and informatics pipelines for the analysis of cancer drug target interactions affected by alternative splicing. Recently, Davuluri has developed novel deep learning approaches for understanding the DNA language, and how genetic and epigenetic changes in the non-coding genome alter the DNA linguistics. Working at the interface of artificial intelligence and genomics, Davuluri is one of the first groups to develop genomic large language model, called “DNABERT”. Released in 2021, DNABERT has been widely used in understanding and decoding genomic and epigenomic languages. The DNABERT model can predict allele-specific activity based only on local nucleotide sequence context, and prioritize candidate transcription-factor-binding sites, core-promoters and splice sites that are sensitive to variants at genome-scale.  Building on DNABERT’s success, his group is developing informatics methods to calculate genome-wide mutational scores, based on whole genome sequence data, and integrate other biomedical data, such as histology and RNA expression to improve prognosis and cancer outcome prediction.