Works on application of statistical methods and modeling to large-scale RNA expression data. In particular, he focuses on inconsistencies between the observed sequence-content of the RNA, as found in deep-sequencing experiments, and the reference genome. These are mostly due to technical and biological noise, but also include signatures of important regulated RNA modifications, called "RNA editing". The crucial role of these RNA modifications to the proper function of the cell and the organism has been increasingly recognized in recent decades, with A-to-I RNA editing being the most common modification. Computational approaches, such as the ones developed and employed by Eisenberg and colleagues, have become the standard tool for detection, quantification and analysis of A-to-I RNA editing events.
Research achievements include: development of several strategies for detection of new editing events in various classes of RNA molecules (repetitive elements, coding sequences, micro-RNAs, etc); quantitative characterization of editing in normal and cancerous human tissues, showing a global deregulation of editing in tumors; reporting the exceptional level of recoding (change of the genomic data content that codes for protein sequences) in cephalopods, showing ~3 orders of magnitude higher level of recoding compared to any other species tested.
Future directions include: (i) study of the role of editing in protection against false activation of the innate immune system; (ii) developing a robust and reliable method for profiling the recoding sites in human and other mammals; (iii) understanding the evolution of recoding events.