We have assembled a large cohort of donors with various head&neck cancers, who have kindly provided matched tumorigenic and healthy adjacent tissue samples. We have profiled a set of immune cells from these samples, namely T cells, which have been extensively shown to play an important role in cancer immunotherapy. Analyses of the healthy adjacent tissue samples are undergoing, and we will be reporting on a comprehensive T-cell atlas of the head&neck compartments. Yet, there remain massive data from the tumor-related side of the project to be analyzed. These data include single-cell RNA-seq, surface protein expression and T-cell receptor (TCR)-seq modalities, all spanning multiple anatomical sites as well as different types of head&neck cancers.

The Amariuta Lab sometimes has bandwidth to take on talented undergraduate students with research or course experience in bioinformatics and statistical data analysis. We have a variety of predefined projects but are also open to student-led projects and ideas that fall within the general scope of research in our lab. These projects generally involve mapping the genetic component of gene expression and, separately, complex traits and polygenic diseases, in order to identify putative disease-critical genes that could serve as predictive biomarkers or therapeutic targets. All projects are in the area of statistical and population genetics, aiming to understand population-specific and shared genetic effects across diverse cell types and tissues, via the integration of high dimensional genomic data with globally diverse DNA sequence (genotyping) and disease data (phenotyping).

The Amariuta Lab is happily accepting PhD rotation students. We have a variety of predefined projects but are also open to student-led projects and ideas that fall within the general scope of research in our lab. One potential rotation project involves the investigation of different ways to estimate gene expression heritability (e.g., the proportion of gene expression variance that can be explained by genetics), including using cutting edge fine-mapping algorithms and single cell RNA-sequencing data. A second potential project involves developing a new method to map the cell-type-specificity of the genetic component of gene expression regulation, which is often confounded by strong patterns of co-expression and co-regulation across genes and other cell types. A third potential project involves mapping the causal genes underlying putative causal tissues and cell types mapped via our previously published method called Tissue Co-regulation SCore regression (TCSC). A fourth potential project involves identifying putative causal tissues and cell types underlying the genetic correlation and hence pleiotropy of immune-mediated diseases. Lastly, we are also beginning to explore deep learning models to use sequence data to predict gene expression levels in novel ways and would welcome students who wish to gain skills in any of these areas.

The implementation of functional precision medicine, which combines genomic profiling with drug sensitivity testing of patient’s tumor cells, has potential to identify personalized and effective treatment options. This project aims to build treatment-response models based on available molecular profiles and patient-matched drug-response data obtained from pediatric brain tumor patients at Rady Children’s Hospital San Diego and at other cancer centers. The goal is to predict novel drug and biomarker pairs to enable better treatment decisions.

In this project, we aim to develop cheap methods for evaluating ecology using shallow sequencing. Methods such as genome skimming have the potential to enhance the study of biodiversity, but new algorithms are needed to enable such studies. That's the goal of this effort. Most methods we develop are kmer-based. The work involves both working with real data and simulations. We work with partners from outside academia as well. The work is shared with Prof. Bafna.

In this broadly defined project, we develop new machine-learning methods to infer or use phylogenies. The goal is to see if machine learning methods can beat existing methods in specific tasks such as phylogenetic placement or outlier detection. The focus is on finding innovative ways to update machine learning methods or adopt them to phylogenetics. We also work on incorporating phylogenies as prior data into machine learning methods for answering other questions (e.g., metagenomic classification).

In this broadly defined project, we develop novel methods for inferring phylogenetic trees from genome-wide datasets. The project involves modeling evolutionary processes, developing scalable algorithms, analyzing their statistical properties in theory, applying them to large simulated and biological data to evaluate them, and providing software to the community.