Leveraging Heterogeneity to Accelerate Translational Medicine

Decoding "Dirty" Data

We develop machine learning methods and tools to detect signals in “dirty” (noisy) data.

Computational Immunology

We analyze immune response as a system using machine learning methods to understand its role in human diseases.

Translational Medicine

Our mission is to leverage “dirty data” and apply machine learning to develop diagnostics and therapies to improve global health.

Khatri Lab in the News

Our Research Philosophy

We are taking advantage of advances in machine learning to turn the traditional paradigm in biomedical research on its head. Instead of limiting heterogeneity in data, as a traditional biomedical research experiment does, we are embracing heterogeneity in data. We firmly believe that biological and technical heterogeneity in data is a blessing in disguise and can accelerate translational medicine. We have repeatedly demonstrated that biological and technical heterogeneity in data, coupled with novel machine learning methods, is not only desired, but required to identify robust disease signatures across patient populations that are diagnostic, prognostic, therapeutic, and mechanistic.

Machine Learning for Multicohort Analysis: Epigenetic, Cellular, and Transcriptional

We take advantage of the biological and technical heterogeneity that exists in diseases, methods that generated the data, and genetically distinct cohorts throughout the world. The results from our “in silico” experiments lead to new hypothesis generation, and “wet lab” validation of the findings. This approach has already been shown to work in sepsis, influenza vaccination and infection, cancer, autoimmune diseases, and transplant rejection.

Khatri Lab Research Highlights

Comparison of Pfizer vaccine with other vaccines

Systems immunology of Pfizer-BioNTech mRNA vaccine (Preprint 2021)

Histone clipping in monocytes to macrophage differentiation

EpiTOF identifies novel mechanism of cellular differentiation (Nature Immunology 2021)

Panvirus analysis of host immune response

Conserved host response to viral infection predicts severity (Immunity 2021)

Xpert HR cartrige

Translation of 3-gene signature for TB diagnosis in a point-of-care cartridge (J Clin Microbio 2021)

Interferon-independent signature of SLE (JCI Insight 2020)

Predicting poor outcome in fibrosis (Lancet Respir Med 2019)

Predicting progression to severe dengue (Cell Reports 2019)

EpiTOF: platform for single-cell epigenome profiling (Cell 2018)

Systems immunology analysis of Mtb infection (Nature 2018)

Increasing accuracy of cellular deconvolution (Nature Communications 2018)

NK cells predict susceptibility to influenza prior to exposure (Genome Medicine 2018)

Predicting response to anti-TNF prior to treatment (Gut 2018)

Predict response to influenza vaccine (Sci Immunology 2017)

Ligand discovery for tumor-derived T cell receptors (Cell 2017)

Framework for leveraging heterogeneity (Nucleic Acids Research 2017)

Distinguishing viral and bacterial infections (Sci Transl Med 2016)

Diagnosis of tuberculosis (Lancet Respir Med 2016)

Host response to viral infections (Immunity 2015)

Diagnosis of sepsis in trauma patients (Sci Transl Med 2015)

Drug repurposing for transplant rejection (J Exp Med 2013)

Heterogeneity is Us

In the Khatri lab, we have created an environment where people with different expertise and interest collaborate to improve our understanding of immune system and accelerate translational medicine.

© Copyright 2019 -    |   Khatri Lab   |   All Rights Reserved