Dr Irina Mohorianu

Irina Mohorianu


Turing Fellow

Partner Institution


Irina’s background in Computer Science was streamlined after she finished her BSc (hons), towards Applied Data Science (Machine Learning and Bioinformatics). During her PhD she studied small non-coding RNAs (sRNAs), first in plants and then in animal systems, too. Soon after completing the first year of post-doc, she was awarded a BBSRC grant on which she was Researcher Co-I. While working on this project she started to integrate mRNA /sRNA expression and became interested in Gene Regulatory Networks (GRNs). She also continued working on developing new methods for the UEA sRNA Workbench. After her post-doc, she was briefly Bioinformatics Lead at the Oxford Vaccine Group (Department of Paediatrics, School of Medical Sciences, Oxford University) and Lecturer in Computer Science at University College, Oxford, teaching Imperative Programming, Algorithms and Machine Learning.

At the Cambridge Stem Cell Institute, she is Head of Bioinformatics & Scientific Computing; she leads a team of Bioinformaticians who develop new Machine Learning & AI methods and also support the CSCI PIs with data analyses. Through her collaborations (both within the University of Cambridge and the community) she aims to understand the intricacies of the underlying GRNs, across time or spatial series and across modalities through analyses of bulk and single cell multi-omics datasets. Integrative analyses are also combined with imaging approaches intertwined with other high throughput measurements. The training of future bioinformaticians, with various backgrounds, is also an important pillar of her group; she is lecturer/ tutor on various modules, with a significant Machine Learning/ Data Science/ Information Theory component.

Research interests

Irina’s research interests span the computational challenges of bioinformatics (high throughput sequencing across various modalities, co-sequencing, imaging linked to spatial transcriptomics and the processing of large medical measurements). A decade ago, the main question was “what is differentially expressed”; current challenges focus on “why is a gene differentially expressed”. The functional characterisation of a gene (or multiple genes, part of a pathway) poses additional computational difficulties due to noise in quantification, biases, and often incomplete set of measurements. While building optimised tools to ensure robust and reproducible analyses, her goal is to discern the relevant patterns that could bring quantitative medicine closer to the goal of personalised medicine.