My research focuses on the development of novel algorithms for predicting the structures and functions of proteins and RNA. The goal is to enable scalable, genome-wide annotation of human and microbial proteins, as well as non-coding RNAs. I have previously developed leading methods for predicting protein structure and function. Notable contributions include my work on COFACTOR, MetaGO, and StarFunc. These were among the highest-performing Gene Ontology prediction methods in recent Critical Assessment of Function Annotation (CAFA) challenges. They notably demonstrated the benefits of using computational structure models for protein function prediction. Additionally, I co-developed TripletRes, a deep learning-based program for predicting inter-residue contact and distance within proteins. I also designed D-QUARK, which enables distance-based protein folding guided by distances derived through deep learning. Both D-QUARK and the TripletRes algorithm stood among top servers in the recent Critical Assessment of Structure Prediction (CASP) 13 and 14 experiments. Beyond that, the DeepMSA package, which I created for constructing multiple sequence alignments, is now a commonly utilized tool by several CASP teams. Leveraging my previous expertise in protein structure and function modeling, my recent research has shifted its focus towards exploring new methods of RNA structure and function analysis. In this domain, I have developed tools such as US-align, the first tool capable of aligning the structures of RNAs, DNAs, and proteins in both monomeric and oligomeric forms via a unified TM-score. My rMSA pipeline, another significant contribution, has gained popularity among top-performing CASP teams. It aids in the generation of RNA sequence alignments necessary for deep learning-based RNA structure prediction. Between 2017 and 2022, I had 47 peer-reviewed manuscripts published, with 12 of them, in high-impact journals such as Nature Methods, Nucleic Acid Research, and Bioinformatics, being first author papers. The manuscript introducing the COFACTOR algorithm received over 400 citations. Another manuscript, which explores the zoonotic origin of SARS-CoV-2 using predicted protein structure and metagenomic sequence assembly, has been cited 300 times.