2 Postdoc positions in Machine Learning for the Microbiome at Harvard Medical School

[Georg Gerber]

Two post-doc positions have recently become available in my lab at HMS. The positions are available now, but the start date is flexible (e.g., summer start date is fine for students graduating in May.) My lab focuses on developing Bayesian ML methods for understanding the microbiome. These post-docs would be a nice opportunity to develop novel ML methods with real bio applications and also to gain microbiome experience. I’m definitely open to students who’ve worked in completely different areas but would be keen to learn about the microbiome.

Machine Learning/Computational Biology Post-doctoral Fellow Opportunities at Harvard Medical School

Two post-doctoral positions available (with flexible start dates) to develop novel machine learning/computational biology approaches to model and understand mammalian microbiomes. Projects include elucidating fundamental rules governing the formation and maintenance of complex microbial ecosystems in the mammalian gut under the National Science Foundation funded project, MTM2: The rules of microbiota colonization of the mammalian gut project (https://gerber.bwh.harvard.edu/mtm-2-the-rules-of-microbiota-colonization-of-the-mammalian-gut) and developing novel computational methods to analyze longitudinal microbiome data under the NIGMS R01 “Bayesian Machine Learning Tools for Analyzing Microbiome Dynamics.” 

The position will give you the opportunity to develop advanced machine learning/computational biology methods while working on real, biologically relevant problems. Techniques we use include Bayesian nonparametric models, dynamical systems inference from sparse data, interpretable models, approximate inference methods and relaxations of discrete variables to enable fully-differentiable models.

The candidate is expected to engage with the broader machine learning and computational biology communities by presenting work at top conferences, as well as publishing applications of new methods in high impact journals. Although some experience modeling biological or other complex systems is required, microbiome specific knowledge is not required. This could be a good fit for either someone with a strong machine learning background who wants to get domain-specific research experience, OR someone with a strong mathematical background who wants to get more machine learning experience. 

About the lab: the Gerber Lab (http://gerber.bwh.harvard.edu) develops novel statistical machine learning models and high-throughput experimental systems to understand the role of the microbiota in human diseases, and applies these findings to develop new diagnostic tests and therapies. A particular focus of the Gerber Lab is understanding dynamic behaviors of host-microbial ecosystems. Our work in this area includes the fully Bayesian MCTIMME, MDSINE and MITRE algorithms, for discovering temporal patterns in microbiome data, inferring dynamical systems models from microbiome time-series data, or predicting host status from microbiome time-series data with human interpretable rules. We have applied these methods to a number of clinically relevant questions including understanding dynamic effects of antibiotics, infections and dietary changes on the microbiome, and designing bacteriotherapies for C. difficile infection and food allergy. We also apply our methods to synthetic biology problems, to engineer consortia of bacteria for diagnostic and therapeutic purposes.

Environment:  the Gerber Lab is located in the Division of Computational Pathology (http://comp-path.bwh.harvard.edu), which Dr. Gerber heads, at Brigham and Women’s Hospital (BWH) at Harvard Medical School (HMS), and the Massachusetts Host-Microbiome Center (http://metagenomics.partners.org), which Dr. Gerber co-directs. BWH, an HMS affiliated teaching hospital is adjacent to the HMS main quad and is the second largest non-university recipient of NIH research funding. The broad mandate of the BWH Division of Computational Pathology is to develop and apply advanced computational methods for furthering the understanding, diagnosis and treatment of human diseases. The Division is situated within the BWH Department of Pathology, which houses over 40+ established investigators, 50+ postdoctoral research fellows, and 100+ research support staff. In addition, BWH is part of the greater Longwood Medical Area in Boston, a rich, stimulating environment conducive to intellectual development and research collaborations, which includes HMS, Harvard School of Public Health, Boston Children’s Hospital and the Dana Farber Cancer Institute.

Qualifications:

- PhD in computer science, computational biology, applied mathematics, statistics, or other quantitative discipline

- Excellent publication track record

- Strong mathematical background with track record developing novel models and methods

- Solid programming skills in Python; this isn’t a software engineering job, but you will need to be able to develop efficient implementations and apply your work to real biological data

- Experience modeling biological or other complex systems required; microbiome experience desirable, but not required

- Superior communication skills and ability to work on multidisciplinary teams

- Ability to reside in the U.S. and legally work in the country; there is an opportunity to work remotely during the COVID-19 pandemic, but funding sources require the post-doc to work in the U.S.

Send cover letter, CV and brief research statement to gge...@bwh.harvard.edu. In your CV, indicate whether you are a U.S. citizen/permanent resident or visa holder (and list visa type.) 

We are an equal opportunity employer and all qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability status, protected veteran status, gender identity, sexual orientation, pregnancy and pregnancy-related conditions or any other characteristic protected by law.