Publications
PhD Thesis
Scaling in the Immune System, PhD Thesis, University of New Mexico, USA, 2013 (pdf) (computational immunology talk part 1) (computational immunology talk part 2) (zenodo link) (OSF project link)
Key Publications
1) Patient and public involvement to build trust in artificial intelligence: a framework, tools and case studies, Soumya Banerjee, Phil Alsop, Linda Jones, Rudolf Cardinal, Patterns 3(6):100506, 2022 (journal link) (resource) (resource) (blog post for general audience)
(Cell Press publishing group)
(highlights: This includes a framework, case studies and tools to involve patients in AI research. Involving patients and carers in research will help build trust in AI.)
2) A class-contrastive human-interpretable machine learning approach to predict mortality in severe mental illness, Soumya Banerjee, Pietro Lio, Peter Jones, Rudolf Cardinal, Nature Partner Journal Schizophrenia, 7, 60, 2021
(Nature Partner Journal, Nature Publishing Group, Impact Factor = 6.3) (link) (journal home) (code) (code)
3) Deconvolution of monocyte responses in inflammatory bowel disease reveals an IL-1 cytokine network that regulates IL-23 in genetic and acquired IL-10 resistance Gut, 2020
(British Medical Journal publishing group, Impact Factor = 19.8) (link) (preprint)
4) Hydroxychloroquine: balancing the needs of LMICs during the COVID-19 pandemic, Soumya Banerjee, Lancet Rheumatology, 2(7):385-386, 2020 (link) (link) (preprint)
(Lancet Publishing Group)
5) Predictive metabolomic profiling of microbial communities using amplicon or metagenomic sequences, Himel Mallick, Eric A. Franzosa, Lauren J. Mclver, Soumya Banerjee, et al., Nature Communications, 10(1):3136, 2019
(Impact factor = 12.2) (journal link) (code)
Selected peer-reviewed conferences and journals (in order of most relevant publications)
NOTE: In my field (Computer Science), people submit to conferences and conferences are peer-reviewed
Here is a link to preprints of my most significant publications (a more complete set of preprints can be found here, here, and here)
Here is a video where I explain my research
25) Generating Complex Explanations for Artificial Intelligence Models: An Application to Clinical Data on Severe Mental Illness, Soumya Banerjee, Life, 14(7), 807, 2024 (journal link) (code) (code)
24) Intelligence in animals, humans and machines: a heliocentric view of intelligence? Halfdan Holm, Soumya Banerjee, AI & Society, 2024 (journal link) (preprint) (preprint) (preprint)
(highlights: A few centuries back we used to think that the entire Universe revolves around Earth. Are we making a similar mistake with AI today? Do we not think that all intelligence revolves around us humans? And that we are the most intelligent species on Earth? Are we not very anthropocentric in our view of intelligence? After all what about intelligence in all its varied forms in animals and plants? We suggest a radical reconceptualisation of intelligence.)
23) Neural networks for abstraction and reasoning: Towards broad generalization in machines, Mikel Bober-Irizar, Soumya Banerjee, 2024 (under review) (preprint) (preprint) (code) (blogpost)
22) Involving patients in artificial intelligence research to build trustworthy systems, Soumya Banerjee, Sarah Griffiths, AI & Society, 2023 (journal link) (preprint) (resource) (blog post for general audience)
(Impact Factor 3.0, part of Springer Nature)
(highlights: I work in the field of AI applied to healthcare. More often than not, patients just end up being data points in our research. Can we deeply involve patients in AI research? Can this help build trust in AI? We propose a responsible AI framework of how to involve patients in research.
Artificial intelligence (AI) is helping researchers and clinicians make great strides in healthcare. However, many people misunderstand AI. Hence, it is critical to engage the public in AI research in healthcare and prove to them that AI applied to healthcare can bring benefits. One way to achieve this is by getting the community involved in co-designing better AI systems in healthcare projects. We believe that integrating patient and public involvement in AI projects may help to adopt and accept these technologies, by making them more transparent. We argue that AI algorithms for healthcare should be co-designed with patients and healthcare workers, so that they are useful and trustworthy.)
21) Patient and public involvement to build trust in artificial intelligence: a framework, tools and case studies, Soumya Banerjee, Phil Alsop, Linda Jones, Rudolf Cardinal, Patterns 3(6):100506, 2022 (journal link) (resource) (resource) (blog post for general audience)
(Cell Press publishing group)
(highlights: This includes a framework, case studies and tools to involve patients in AI research. Involving patients and carers in research will help build trust in AI.)
20) Software Application Profile: ShinyDataSHIELD—an R Shiny application to perform federated non-disclosive data analysis in multicohort studies, Xavier Escribà-Montagut, Yannick Marcon, Demetris Avraam, Soumya Banerjee, Tom R P Bishop, Paul Burton, Juan R González, International Journal of Epidemiology, dyac201, 2022
(link) (code) (demo) (tutorial)
(Oxford University Publishing Group, Impact Factor 9.8)
(highlights: This is an accessible user interface for statistical and machine learning. The interface allows non-technical users to carry out analysis on data using a federated privacy preserving platform.)
19) dsSurvival 2.0: Privacy enhancing survival curves for survival models in the federated DataSHIELD analysis system, Soumya Banerjee, Tom Bishop, BMC Research Notes 16, 98, 2023
(link)(preprint)(lay summary)(code)(code)(code)(code)
(highlights: a package and tools for privacy preserving survival curve visualization in clinical informatics)
18) dsSynthetic: Synthetic data generation for the DataSHIELD federated analysis system, Soumya Banerjee, Tom Bishop, BMC Research Notes 15(1):230, 2022
(link) (code, code, code) (preprint) (supplementary material)
(Springer Nature publishing group)
(highlights: a package for generating synthetic data for privacy preserving analysis in healthcare)
17) dsSurvival: Privacy preserving survival models for federated individual patient meta-analysis in DataSHIELD, Soumya Banerjee, Ghislain Sofack, Thosoris Papakonstantinou, Demetris Avraam, Paul Burton et al., BMC Research Notes, 15(1):197, 2022 (paper) (code) (code) (code) (tutorial) (preprint)
(Springer Nature publishing group)
(highlights: a package and tools for privacy preserving survival analysis in clinical informatics)
16) A class-contrastive human-interpretable machine learning approach to predict mortality in severe mental illness, Soumya Banerjee, Pietro Lio, Peter Jones, Rudolf Cardinal, Nature Partner Journal Schizophrenia, 7, 60, 2021
(Nature Partner Journal, Nature Publishing Group, Impact Factor = 6.3) (link) (journal home) (code)
15) Simulating a community mental health service during the COVID-19 pandemic: effects of clinician-clinician encounters, clinician-patient-family encounters, symptom-triggered protective behaviour, and household clustering, Frontiers in Psychiatry, 12, 196, 2021
(Impact Factor = 3.5) (link) (preprint) (code)
14) Optogenetic tuning reveals Rho amplification-dependent dynamics of a cell contraction signal network, Dominic Kamps, Johannes Koch, Victor O. Juma, Eduard Campillo-Funollet, Melanie Graessl, Soumya Banerjee, Tomáš Mazel, Xi Chen, Yao-Wen Wu, Stephanie Portet, Anotida Madzvamuse, Perihan Nalbant, Leif Dehmelt, Cell Reports, 33(9):108467, 2020
(Cell Press publishing group, Impact Factor = 8.1) (link) (general summary) (archived general summary) (general summary OSF) (code) (code)
13) Deconvolution of monocyte responses in inflammatory bowel disease reveals an IL-1 cytokine network that regulates IL-23 in genetic and acquired IL-10 resistance Gut, 2020
(British Medical Journal publishing group, Impact Factor = 19.8) (link) (preprint)
12) Hydroxychloroquine: balancing the needs of LMICs during the COVID-19 pandemic, Soumya Banerjee, Lancet Rheumatology, 2(7):385-386, 2020 (link) (link) (preprint)
(Lancet Publishing Group)
11) Predictive metabolomic profiling of microbial communities using amplicon or metagenomic sequences, Himel Mallick, Eric A. Franzosa, Lauren J. Mclver, Soumya Banerjee, et al., Nature Communications, 10(1):3136, 2019
(Impact factor = 12.2) (journal link) (code)
10) The early impact of COVID-19 on mental health and community physical health services and their patients’ mortality in Cambridgeshire and Peterborough, UK Journal of Psychiatric Research, 131, 244-254, 2020
(Impact factor = 4.4) (link)
9) Influence of correlated antigen presentation on T cell negative selection in the thymus, Soumya Banerjee, SJ Chapman, Journal of the Royal Society Interface, 15(148), 20180311, 2018
(Impact Factor = 4.3) (link) (link) (journal link) (main and supplementary material combined) (media summary) (code)
8) Modelling the effects of phylogeny and body size on within-host pathogen replication and immune response, Soumya Banerjee, Alan Perelson, Melanie Moses, Journal of the Royal Society Interface 14(136), 20170479, 2017
(Impact Factor = 4.3) (preprint PDF) (link) (supplementary materials) (talk part 1) (talk part 2)
(media summary) (media coverage) (coverage)
(highlights: We combine data on infectious diseases from different species using machine learning. We propose a new competency metric that allows us to link within-host viral dynamics to between-host spread of diseases. Our technique can be applied to other emerging diseases like Zika and Ebola virus that infect multiple species.)
7) An excitable Rho GTPase signaling network generates dynamic subcellular contraction patterns, Melanie Graessl, Johannes Koch, Abram Calderon, Dominic Kamps, Soumya Banerjee, Tomáš Mazel, Nina Schulze, Jana Kathrin Jungkurth, Rutuja Patwardhan, Djamschid Solouk, Nico Hampe, Bernd Hoffmann, Leif Dehmelt, Perihan Nalbant. Journal of Cell Biology 216(12), 4271-4285, 2017
(Impact Factor = 9.8) (link) (preprint) (work reviewed and discussed here) (link to a movie) (code and model file for intra-cellular network link link )
(highlighted in Journal of Cell Biology spotlight) (summary for general audience)
6) Estimating biologically relevant parameters under uncertainty for experimental within-host murine West Nile virus infection, Soumya Banerjee, Jeremie Guedj, Ruy Ribeiro, Melanie Moses & Alan Perelson Journal of the Royal Society Interface, 13(117), 20160130, 2016
(Impact Factor = 4.3) (preprint) (link) (supplementary material) (summary for general audience) (summary) (coverage)
(highlights: West Nile virus (WNV) causes viral encephalitis in humans, and is related to viruses such as Dengue and Zika that are also of significant public health concern. We have developed a computational method to determine characteristics of WNV infection even in the face of limited experimental data. This could be applicable to other emerging diseases like Zika virus for which there is little data. It may be particularly useful to estimate the potential rate of within-host viral reproduction early in an outbreak in order to assess the epidemic potential of emerging pathogens.)
5) Competitive dynamics between criminals and law enforcement explains the super-linear scaling of crime in cities, Soumya Banerjee, Manuel Cebrian, Pascal van Hentenryck, Palgrave Communications 1, 15022, 2015
(Nature Publishing Group) (paper) (preprint) (data) (supplementary information) (coverage) (code)
(top 5% of all papers covered by Altmetric as of February 2016)
(highlights: Larger cities have disproportionately more crime per capita compared to smaller cities [super-linear scaling of crime]. We used techniques from dynamical systems and complex systems to explain the super-linear scaling of crime in cities and other socio-technological systems.)
4) A bioorthogonal small-molecule switch system for controlling protein function in cells, Peng Liu, Abram Calderon, Georgios Konstantinidis, Jian Hou, Stephanie Voss, Xi Chen, Fu Li, Soumya Banerjee, Jan‐Erik Hoffmann, Christiane Theiss, Leif Dehmelt, Yao‐Wen Wu (2014) 53(38), 10049-10055, Angewandte Chemie
(Impact Factor = 13.7) (preprint) (link)
(highlights: a patented technique [International patent application PCT/EP2013/060890] and the first reversible small-molecule system for controlled protein interaction in live cells. I used statistical techniques to analyze the data and performed automated cell tracking using ImageJ and CellProfiler)
3) Science and technology consortia in US biomedical research: A paradigm shift in response to unsustainable academic growth, Curt Balch, Hugo Arias-Pulido, Soumya Banerjee, Alex K. Lancaster, Kevin B. Clark, Michael Perilstein, Brian Hawkins, John Rhodes, Piotr Sliz, Jon Wilkins and Thomas W. Chittenden, (2014) BioEssays
(Impact Factor = 5.4) (link) (pdf)
2) A spatial model of the efficiency of T cell search in the influenza-infected lung, Drew Levin, Stephanie Forrest, Soumya Banerjee, Candice Clay, Judy Cannon, Melanie Moses and Frederik Koster, Journal of Theoretical Biology, 398(7), 52-63, 2016
(Impact Factor = 2.3) (journal link) (pdf) (supplementary section) (Supplementary videos of agent based model) (code)
1) Scale Invariance of Immune System Response Rates and Times: Perspectives on Immune System Architecture and Implications for Artificial Immune Systems, Soumya Banerjee and Melanie Moses, Swarm Intelligence, 4(4), 301-318, DOI: 10.1007/s11721-010-0048-2, 2010
(Impact Factor = 2.2) (journal link) (preprint) (pdf) (arXiv) (bibTeX) (code) (code)
Other manuscripts (in progress, under review and miscellaneous)
16) A human-interpretable machine learning approach to predict mortality in severe mental illness (in review) (preprint)
(highlights: Interpretable machine learning applied to healthcare)
17) ShinyDataSHIELD: An R Shiny application to perform federated non-disclosive data analysis in multi-cohort studies (in review) (code)
(highlights: a user-interface and package for privacy preserving clinical informatics)
18) dsSurvival: Privacy preserving survival models for federated individual patient meta-analysis of clinical data (code) (code) (code) (tutorial) (preprint)
(highlights: a package and tools for privacy preserving survival analysis in clinical informatics)
19) SAFEST: A Safeguarding Analytical Framework for dEcentralised Sensitive daTa, Patricia Ryser-Welch, Leire Abarrantegui, Soumya Banerjee International Journal of Advancements in Technology, 13:213, 2022
(highlights: review of privacy preserving techniques in healthcare)
20) Disclosure control in DataSHIELD
(highlights: review of privacy preserving techniques in the DataSHIELD federated data analysis platform)
21) Generating complex explanations from machine learning models using class-contrastive reasoning. Yujia Yang, Soumya Banerjee, 2023
(highlights: a new explainable AI algorithm to generate complex explanations/stories from data)
22) Data-centric explainability and generating complex stories as explanations from machine learning models Yujia Yang, Soumya Banerjee, 2023
(highlights: a new data-centric explainable AI algorithm to explain models)
23) Generating complex explanations for artificial intelligence models: an application to clinical data on severe mental illness, Soumya Banerjee, 2024 (preprint)(preprint) (code) (code)
(highlights: a new data-centric explainable AI algorithm to explain models applied to data on mental health)
24) Enhancing patient stratification and interpretability through class-contrastive and feature attribution techniques, Sharday Olowu, Neil Lawrence, Soumya Banerjee (preprint) (code)
Talks
1) "Immune System Inspired Strategies for Distributed Systems", S. Banerjee & M. Moses. 6th Annual Computer Science at UNM Student Conference (CSUSC) 2010 (slides) (talk) (login for accessing talk: username - video, password - speaker0watch)
2) "Modular RADAR: An Immune System Inspired Search and Response Strategy for Distributed Systems". The 9th International Conference on Artificial Immune Systems (ICARIS), 2010, Edinburgh, UK (slides)
3) Playlist of my talks (link)
Awards
1) 2010 UNM Student Award for Innovation in Informatics (the award is given to a UNM graduate or undergraduate student for the best paper describing innovation or research in the field of biomedical informatics)
Research News Coverage
1) Research covered by the University of Oxford (pdf, link)
2) Department of Computer Science, University of New Mexico (featured on department webpage) (pdf)
3) School of Engineering, University of New Mexico (featured on school webpage) (pdf)
4) School of Health Sciences, University of New Mexico (pdf)
