A Community Challenge for Inferring Genetic Predictors of Gene Essentialities through Analysis of a Functional Screen of Cancer Cell Lines

Abstract

SUMMARY We report the results of a DREAM challenge designed to predict relative genetic essentialities based on a novel dataset testing 98,000 shRNAs against 149 molecularly characterized cancer cell lines. We analyzed the results of over 3,000 submissions over a period of 4 months. We found that algorithms combining essentiality data across multiple genes demonstrated increased accuracy; gene expression was the most informative molecular data type; the identity of the gene being predicted was far more important than the modeling strategy; well-predicted genes and selected molecular features showed enrichment in functional categories; and frequently selected expression features correlated with survival in primary tumors. This study establishes benchmarks for gene essentiality prediction, presents a community resource for future comparison with this benchmark, and provides insights into factors influencing the ability to predict gene essentiality from functional genetic screens. This study also demonstrates the value of releasing pre-publication data publicly to engage the community in an open research collaboration.

Signatures of cell death and proliferation in perturbation transcriptomics data - from confounding factor to effective prediction

“…While baseline models -using perturbation and cell line information -were also able to reach effective performance within a given dataset (CTRP or Achilles), this effective across dataset prediction is unique for the perturbation signature based models. Also several studies investigated the predictability of drug sensitivity (Ali and Aittokallio, 2018;Costello et al, 2014;Iorio et al, 2016) and gene essentiality (Gönen et al, 2017) with good performance, but translatable prediction was not attempted between these different, but related phenotypes. There could be two main reasons for the effective across dataset prediction performance of our methods.…”

Highly Accurate Cancer Phenotype Prediction with AKLIMATE, a Stacked Kernel Learner Integrating Multimodal Genomic Data and Pathway Knowledge

“…We chose these 37 tasks (out of 5711 available consensus profiles) because they had at least 10 cell lines showing strong (>2 sd from mean) knockdown viability response and are in the top quartile by variance of all consensus profiles (see Supplementary Information). Based on the results of the DREAM9 gene essentiality prediction challenge [29], we expected this task to be significantly harder than the other case studies.…”

“…We compared AKLIMATE's performance to three of the five top performing methods in DREAM9 [29] as well as three baseline algorithms. We briefly describe the DREAM9 subchallenge 1 top performers next.…”

“…Both methods were run with the same hyperparameters and pathway collections as in DREAM9 [29] [78]. Kernelized Gaussian Process Regression (GPR) took 3rd place -it used extensive filtering steps to reduce the input feature dimensionality, followed by principal component analysis, and finally Gaussian Process regression with covariance computed from the principal components [29] (see also https://www.synapse.org/#!Synapse:syn2664852/wiki/68499 for implementation and model description). We downloaded the code from the Synapse URL and ran it with the published DREAM9 hyperparameters.…”

“…To prevent overfitting, MKL methods generally include a regularization term -e.g., L 1 sparsity-inducing norm on the kernel weights [59] or the elastic net [74]. Prior knowledge can also be integrated by constructing individual kernels from a pathway's member features within each data type [29] [27] [66] [78] [50]. Indeed, MKL methods with prior knowledge integration [29] [14] have won several Dialogue on Reverse-Engineering Assessment and Methods (DREAM) [70] challenges, including a predecessor of the approach described here [78] [29].…”