Training Program

2019-20 Cancer Systems Biology Trainees

Alex Wenzel, a graduate student in Jill Mesirov's lab (UC San Diego). Gene Set Enrichment Analysis (GSEA) has become essential in the effort to understand molecular pathway behavior in cancer. This approach, and similar methods, are heavily dependent on supporting resources like the Molecular Signatures Database (MSigDB), which provides sets of genes corresponding to biological processes and pathways with which to probe genomic data. Enrichment analyses are only as good as the sets used in their application. In this project, we will devise algorithmic, data-driven approach to refining the gene sets in MSigDB to provide more coherent and context-specific sets and enable more powerful enrichment analyses.

Daniel Schwarz, a graduate student in Jason Gestwicki's lab (UCSF). Protein phosphorylation is a fundamental and essential cellular signaling mechanism. Phosphorylation events that occur on a Ser or Thr preceding a Pro residue, termed Pro-directed phosphorylation, are enriched for cell-cycle progression, apoptosis and differentiation pathways. These phosphorylations, like all peptidyl-prolyl bonds, may exist in distinct cis or trans conformation with rotation occurring at the peptide bond. The evolutionarily conserved peptidyl-prolyl isomerase (PPIase) Pin1 is the only known enzyme to catalyze this isomerization. I will employ an epistatic genome-wide CRISPRi survival screen to identify sensitizing and protective gene knockdowns to Pin1 inhibition in search for genes and biological processes which are differentially toxic in the presence of a Pin1-inhibitor that our group has previously reported.

Jackie Einstein, a graduate student in Gene Yeo's lab (UC San Diego). Tumor progression is a multistage process driven by growth promoting pathways that become upregulated due to amplified or mutated proto-oncogenes. The field has recently embraced the importance of identifying genes functioning upstream and downstream of prominent oncogenes to classify additional key hubs, since scale-free networks are more sensitive to therapeutic disruptions at major hubs. Many studies have suggested that disrupting particular splicing factors, which typically are RBPs, may present an avenue for new therapeutic approaches. I therefore hypothesize that a set of RBPs modulate the pathways required for cancer cells to maintain their gene expression programs. Since individual proteins rarely mediate cellular behaviors, I propose to employ a systems biology approach to reveal the ectopic RNA processing pathways using large scale-functional studies.

Jisoo Park, a postdoctoral fellow in Trey Ideker's lab (UC San Diego). This project will build on ongoing efforts in the CCMI. Using  the cancer pathway structural hierarchy generated in the Ideker Lab as part of Project 3, I plan to add predictive power to those hierarchies by fitting it with functions that translate a patient genotype and therapeutic regimen into a predicted phenotypic response. This deep neural network model, we will call Human DCell (HDCell), will be trained to predict the growth rate of cancer cells in response to perturbations caused by combinatorial genetic disruptions and drugs. As for this training data, we will be greatly aided by the genetic interaction data systematically measured by CCMI members in Project 2 and elsewhere, which will provide the cellular growth rate for many thousands of different combinations of gene disruptions. Model performance will be assessed by cross-validation and by further evaluation of genotypic perturbations not used for model training.

Kirti Chahal, a graduate student in Irina Kufareva's lab (UC San Diego). Aberrant Hedgehog (Hh) signaling pathway activation is found in several cancers, including basal cell carcinoma (BCC), medulloblastoma (MB), pancreatic adenocarcinoma, glioblastoma multiforme (GBM), and others – accounting for ~25% of all cancer deaths. The Hh pathway is activated by binding of sonic hedgehog (Shh) protein to the 12 transmembrane Patched (PTCH) receptor to relieve its suppression of Smoothened (SMO), a 7 transmembrane receptor from the Frizzled family (family F). Hh pathway activation involves trafficking of SMO and several signaling components between the plasma membrane, intracellular membranes, primary cilia (PC) and the nucleus; however, our understanding of the molecular mechanisms behind these trafficking events is limited. The goal of this project is to uncover the SMO interactome, using a new spatiotemporally resolved proteomics technique called APEX and based on proximity biotinylation. The identified proteins will be computationally annotated and ranked by relevance and impact.

Michelle Moritz, a Research Biochemist in Nevan Krogan's and David Agard's lab (UCSF). The goal of this project is to use the state-of-the-art mass spectrometry and cryo-EM facilities in the Krogan and Agard labs to take an innovative and integrative approach to understanding protein complexes important in cancer at a level of atomic detail rarely possible previously. Structural information determined by XL-MS and cryo-EM in conjunction with known crystal structures will be used to generate working models that can inform future experiments to disrupt or manipulate molecular interactions potentially important in cancer. This work will specifically seek to gain an atomic-level understanding of the interaction of the regulatory casein kinase-1δ and the microtubule-nucleating γ-Tubulin Ring Complex (γTuRC) using cross-linking mass spectrometry (XL-MS) and cryo-electron microscopy (cryoEM). This work will serve as a test case for solving structures of other cancer-related protein complexes that were identified through affinity-purification mass spectrometry (AP-MS).