Computational Genomics

Massachusetts Institute of Technology
Computer Science and Artificial Intelligence Laboratory


Prof. David K. Gifford

We study stem cell based developmental biology with original computational methods that build predictive models from high-throughput experiments. We design these experiments with our collaborators in other laboratories to reveal key events during development, including dysfunctions that can lead to human disease. In addition, we are interested in the genetic foundations of human disease, and study the broad question of how an individual’s genotype influences their phenotype.

Current focus area of our laboratory include:

Motor Neuron Development and Disease (

Our laboratory leads an interdisciplinary project that seeks to build computational models of the transcriptional regulatory networks that control the differentiation of neural cells. Elucidating these regulatory networks will enable us to define the regulatory processes that determine a cell's progress to its terminally differentiated state, and understand developmental defects that cause debilitating human diseases such as Spinal Muscular Atrophy. We develop new computational methods for elucidating transcriptional regulatory networks based on the integration of diverse high-throughput experimental data (genome sequence, chromatin structure, transcription factor-DNA binding, gene expression). These methods provide a powerful foundation for discovering the regulatory network control that controls cell differentiation during development.

Pancreatic Development (

We have developed engineered mouse stem cell lines and computational models of pancreatic development to gain insight into potential therapeutics for diabetes. Our stem cell work is identifying in vitro differentiation protocols to create pancreatic progenitors, and we are experimentally elucidating the molecular events that occur during the development of these progenitors using a variety of high-throughput technologies (RNA Seq, ChIP Seq, Mass Spectrometry). Data from these experiments are processed with computational methods developed by our laboratory to reveal biological mechanisms for further exploration.

The Genotype to Phenotype Problem

Working with other laboratories we have discovered that different individuals of the same species can require different sets of genes for survival. Genes that are differentially required for survival are called conditional essential genes. Our work uses a yeast model system that permits us to identify the genetic suppressors that permit one strain to survive without a gene that is necessary for the survival of another strain. Ultimately we aim to elucidate a computational description of the genetic variants that produce a common phenotype using new approaches that reveal complex genetic interactions.