Postdoc research scholar for machine learning in single-cell cancer genomics

Institution/Company:
McGill University
Location:
Montreal, QC, Canada
Job Type:
  • Postdoctoral
Degree Level Required:
Bachelor's, PhD
Apply Now

Postdoc research scholar for machine learning in single-cell cancer genomics

Background

Early detection of cancer before metastases are formed would greatly aid clinical intervention. Genome-wide RNA expression profiling by RNA sequencing (RNA-seq) of cancer driver genes and prognostic biomarkers hold great promise to screen patients for cancer and future cancer recurrence after surgical tumor removal. The advent of single-cell RNA sequencing (scRNA-seq) has unlocked the cell transcriptome at cellular-level resolution, providing theoretically optimal resolution potential. Extensive single cell surveys of tissues in mice or humans in recent studies provide a portrait of variation in gene expression of cell-types in different tissue contexts.

Compared to bulk RNA-seq, however, scRNA-seq is a much newer technology with more unknown technical biases, is much more expensive, is more challenging to perform, and is less sensitive to detect lowly expressed genes and prone to sequencing error. On the other hand, bulk RNA-seq technology has been the primary workhorse in many large-scale transcriptomic studies. In particular, The Cancer Genome Atlas (TCGA) is a repository of multi-omic bulk genome-wide profiles including bulk RNA-seq for over 10,000 tumor samples across 33 cancer types. Despite many publications that stemmed from the TCGA consortium, the reproducibility of many novel cancer marker genes is still in question.

Currently, we lack a probabilistic model that can accurately dissect the cell-type heterogeneity of bulk expression samples by leveraging reference single-cell molecular profiling data as reference. To identify clinically reliable cell-type-specific cancer driver genes with translational medicine in mind, we will develop a hierarchical Bayesian models to deconvolve the 10,000 TCGA bulk RNA-seq profiles over 33 tumors into cell-type-specific gene expression by leveraging the reference scRNA-seq data from publicly available single-cell mouse or human atlases and in-house data from our collaborators. Our approach will not only deconvolve the bulk RNA-seq in tumor samples into cell-type-specific gene expression corresponding to the known cell-types in the reference scRNA-seq panel but will also infer perturbed and new cell-type-specific gene expression discovered in the bulk RNA-seq data that are not observed in the reference scRNA-seq dataset.

Responsibility and Salary

We have budgeted for one postdoctoral scholar (Base salary $60,000 plus benefits) for two years of the proposed research in single-cell multi-omics on cancer. The job employment can start right way. Ideally, the postdoctoral scholar will have strong fundamentals probabilistic models, variational Bayesian, and hands-on experience in single-cell analysis. Backgrounds in neural networks are also desired. The postdoc will be primarily working with Dr. Yue Li, who is the NPI on this grant, along with senior clinical scientists who have access to the cancer patient samples.

Please do not hesitate to contact us if you are interested in this opportunity at yueli@cs.mcgill.ca

– Yue Li Assistant Professor School of Computer Science, McGill University Associate member of Quantitative Life Science, McGill University Associate member of Montreal Institute for Learning Algorithms (MILA) www.cs.mcgill.ca/~yueli/

Responsibilities:

The postdoc will conduct research in consultation with Dr. Yue Li, who is the NPI on this grant and work with senior clinical scientists who have access the cancer patient samples. Detailed responsibilities are listed as follows:

  1. data processing and single-cell analysis
  2. application of existing machine learning models developed by Dr. Li
  3. develop novel Bayesian learning model (working alongside with Dr. Li)
  4. bioinformatics data analysis of multi-omic data
  5. manuscript preparation

Qualifications:

Ideally, the postdoctoral scholar will have strong fundamentals probabilistic models, variational Bayesian, and hands-on experience in single-cell analysis. Backgrounds in neural networks are also desired. The postdoc will be primarily working with Dr. Yue Li, who is the NPI on this grant senior clinical scientists who have access the cancer patient samples.

  • How to Apply

    To apply for this position, please email your CV and/or resume along with the top 2 publications of yours to yueli@cs.mcgill.ca