Research & Activities

We are passionate about developing new algorithms, machine learning and deep learning methods, and their applications to genomics, metagenomics and cancer research. Some of the current projects in the lab include characterizing human genomes and metagenomes sequenced by exciting new technologies, quantifying cancer evolution, study of tumor heterogeneity using genomics and digital pathology images.
View Research & Other Activities →

Structural Variation

Developing algorithms for characterizing structural variations focused on complex and repetitive elements, rare variants and clinical relevant variants.

Tumor Heterogeneity

Quantifying cancer evolution in multiple samples and reconstruction of tumor lineage trees using novel computational methods. Also, building deep learning based classifiers to discriminate heterogeneous sample, using imaging data.


Advancing computational methods in metagenomics for the purpose of deconvolving mixtures, discovering variants and de novo assembly.

Most recent manuscripts

6) Danko, Meleshko, Bezdan, Mason, Hajirasouliha. Minerva: An Alignment and Reference Free Approach to Deconvolve Linked-Reads for Mietagenomics (bioRxiv 217869)
5) Khosravi, Kazemi, Imielinski, Elemento*, Hajirasouliha*. Deep Convolutional Neural Networks Enable Discrimination of Heterogeneous Digital Pathology Images EBioMedicine
4) Toosi, Moieni, Hajirasouliha. BAMSE: Bayesian model selection for tumor phylogeny inference among multiple samples (accepted in ICCABS 2017 and BMC)
3) Ricketts, Popic, Toosi, Hajirasouliha. Using LICHeE and BAMSE for reconstructing cancer phylogenetic trees (In press. Current Protocols in Bioinformatics)
2) Malikic, Ciccolella, Ricketts, Hach, Haghshenas, Rahman, Rashidi Mehrabadi, Seidman, Hajirasouliha*, Sahinalp*. SCePhy - A Constraint Satisfaction Approach for Tumor Phylogeny Reconstruction via Integrative use of Single Cell and Bulk Sequencing Data
1) Meleshko, Mohimani, Hajirasouliha, Medema, Korobeynikov, Pevzner. Reconstructing Biosynthetic Gene Clusters From Assembly Graphs


Below you can see some highlights of our published work. Please also check publications page →

1000 Genomes Project

A catalog of human genome variations in population-scale.


Various methods in Bioinformatics and/or ISMB conferences. Algorithms for Next Generation Sequencing, Structural Variation discovery, tumor heterogeneity and Protein-protein intraction prediction.

Handling Multiple Sequenced Genomes

The CommonLAW package presented on the cover of Genome Research introduces novel combinatorial formulations and algorithms for structural variation discovery among a number of sequenced donor genomes, with the help of a complete reference genome. CommonLAW significantly reduces the false positive rate in detecting structural variation events when compared with conventional methods. (Cover illustration by Azalia Musa, modified by Andres Wanner and Iman Hajirasouliha.)

Want to discuss a project?

We are always keen to establish new collaborations with computational scientists and experimentalists.

Contact us