keshav-motwani.github.io

Keshav Motwani

Keshav Motwani headshot

I am a second-year PhD student in biostatistics at the University of Washington, funded by an NSF Graduate Research Fellowship.

My research interests are broadly in high-dimensional statistical learning and inference motivated by applications in genomics.

I am currently working with Ali Shojaie and Eardi Lila on computationally efficient methods to estimate covariance components in multivariate mixed models. I am also working with Daniela Witten on inference after using a machine learning model to predict the outcome variable of interest and data thinning, an alternative to sample splitting in unsupervised settings.

I did my undergraduate degree at the University of Florida in mathematics and statistics. I worked with Aaron Molstad and Rhonda Bacher on high-dimensional multinomial regression with multiresolution/hierarchical categories. I also worked with Todd Brusko and Victor Greiff on analyzing and developing software for immune receptor sequencing and single-cell genomics data in the context of type 1 diabetes. 

You can see some of my work on my Google Scholar profile and my Github page.

My CV is available here.

Preprints and Submitted Manuscripts

Methods

  1. Keshav Motwani and Daniela Witten. Valid inference after prediction. arXiv, 2023 Paper, Code
  2. Ameer Dharamshi, Anna Neufeld, Keshav Motwani, Lucy L Gao, Daniela Witten, and Jacob Bien. Generalized data thinning using sufficient statistics. arXiv, 2023 Paper

Publications

Methods

  1. Aaron J Molstad and Keshav Motwani. Multiresolution categorical regression for interpretable cell type annotation. In press at Biometrics, 2023 Paper, Software
  2. Keshav Motwani, Rhonda Bacher, and Aaron J Molstad. Binned multinomial logistic regression for integrative cell type annotation. In press at Annals of Applied Statistics, 2023 Paper, Software

Applied

  1. Melanie R Shapiro, Xiaoru Dong, Daniel Perry, James M McNichols, Puchong Thirawatananond, Amanda L Posgai, Leeana Peters, Keshav Motwani, Richard Musca, Andrew Muir, and others. Human immune phenotyping reveals accelerated aging in type 1 diabetes. JCI Insight, 2023 Paper
  2. Chakravarthi Kanduri, Milena Pavlovic, Lonneke Scheffer, Keshav Motwani, Maria Chernigovskaya, Victor Greiff, and Geir K Sandve. Profiling the baseline performance and limits of machine learning models for adaptive immune receptor repertoire classification. GigaScience, 11, 2022 Paper
  3. Milena Pavlovic, Lonneke Scheffer, Keshav Motwani, Chakravarthi Kanduri, Rad- mila Kompova, Nikolay Vazov, Knut Waagan, Fabian LM Bernal, Alexandre Almeida Costa, Brian Corrie, and others. immuneML: an ecosystem for machine learning analysis of adaptive immune receptor repertoires. Nature Machine Intelligence, 2021 Paper, Software
  4. Peter S Linsley, Fariba Barahmand-pour Whitman, Elisa Balmas, Hannah A DeBerg, Kaitlin J Flynn, Alex K Hu, Mario G Rosasco, Janice Chen, Colin ORourke, Elisavet Serti, Vivian H Gersuk, Keshav Motwani, and others. Autoreactive T cell receptors with shared germline-like α chains in type 1 diabetes. JCI Insight, 2021 Paper
  5. Keshav Motwani, Leeana D Peters, Willem H Vliegen, Ahmed Gomaa El-Sayed, Howard R Seay, M Cecilia Lopez, Henry V Baker, Amanda L Posgai, Maigan A Brusko, Daniel J Perry, and others. Human regulatory T cells from umbilical cord blood display increased repertoire diversity and lineage stability relative to adult peripheral blood. Frontiers in Immunology, 11:611, 2020 Paper, Code
  6. Emmi-Leena Ihantola, Henna Ilmonen, Anssi Kailaanmaki, Marja Rytkonen-Nissinen, Aurelien Azam, Bernard Maillere, Cecilia S Lindestam Arlehamn, Alessandro Sette, Keshav Motwani, Howard R Seay, and others. Characterization of proinsulin T cell epitopes restricted by type 1 diabetes–associated HLA class II molecules. The Journal of Immunology, 204(9):2349–2359, 2020 Paper
  7. Mohsen Khosravi-Maharlooei, Aleksandar Obradovic, Aditya Misra, Keshav Motwani, Markus Holzl, Howard R Seay, Susan DeWolf, Grace Nauman, Nichole Danzl, Haowei Li, and others. Cross-reactive public TCR sequences undergo positive selection in the human thymic repertoire. The Journal of Clinical Investigation, 129(6):2446–2462, 2019 Paper