Computational Approaches to Identify Genetic Interactions for Cancer Therapeutics

Benstead-Hume, Graeme; Wooller, Sarah K.; Pearl, Frances M. G.

doi:10.1515/jib-2017-0027

Cited by 5 publications

(6 citation statements)

References 97 publications

(89 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, such direct comparisons pose some challenges: First, the same performance metrics must be used for the comparisons to be valid. Unfortunately, most of the previously published models employ AUC-ROC ( Al-Aamri et al , 2019 ; Benstead-Hume et al , 2017 ; Gabriel del Rio, 2009 ; Mistry et al , 2017 ; Wong et al , 2004 ; Wu et al , 2014 ) to assess model performance, which is not an appropriate metric for imbalanced datasets ( Davis and Goadrich, 2006 ; Saito and Rehmsmeier, 2015 ). Second, the majority of single mutant fitness models were not evaluated in a training/testing setup since they make very strong and restrictive assumptions about the relationship between input features and output phenotype (i.e.…”

Section: Resultsmentioning

confidence: 99%

“…Computational and mathematical models are becoming an essential component for analyzing large biological data sets, such as the one generated by genomics, since they enable the simulation of thousands of manipulations, thereby reducing the number of required laboratory experiments to a more manageable set of key validations. Indeed, significant effort has been invested in developing models that can either predict essential genes, characterized by a lethal phenotype upon deletion ( Campos et al , 2019 ; Campos et al , 2020 ; Cheng et al , 2014 ; Gabriel del Rio, 2009 ; Li et al , 2012 ; Luo and Wu, 2015 ; Zhang et al , 2016 ), or identify novel negative GIs ( Al-Aamri et al , 2019 ; Benstead-Hume et al , 2017 ; Benstead-Hume et al , 2019 ; Chipman and Singh, 2009 ; Paladugu et al , 2008 ; Srivas et al , 2016 ; Wong et al , 2004 ; Young and Marcotte, 2017 ; Yu et al , 2016 ). However, analysis of the outputs of many existing models reveals some limitations (for review, see Madhukar et al , 2015 ).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting and explaining the impact of genetic disruptions and interactions on organismal viability

2022

View full text Add to dashboard Cite

Motivation Existing computational models can predict single- and double-mutant fitness but they do have limitations. First, they are often tested via evaluation metrics that are inappropriate for imbalanced datasets. Second, all of them only predict a binary outcome (viable or not, and negatively interacting or not). Third, most are uninterpretable black box machine learning models. Results Budding yeast datasets were used to develop high performance Multinomial Regression (MN) models capable of predicting the impact of single, double, and triple genetic disruptions on viability. These models are interpretable and give realistic non-binary predictions and can predict negative genetic interactions in triple-gene knockouts. They are based on a limited set of gene features, and their predictions are influenced by the probability of target gene participating in molecular complexes or pathways. Furthermore, the MN models have utility in other organisms such as fission yeast, fruit flies, and humans, with the single gene fitness MN model being able to distinguish essential genes necessary for cell autonomous viability from those required for multicellular survival. Finally, our models exceed the performance of previous models, without sacrificing interpretability. Availability All code used to generate results and figures in this manuscript are available at our Github repository at https://github.com/KISRDevelopment/cell_viability_paper. The repository also contains a link to the genetic interaction (GI) prediction website that lets users search for GIs using the MN models. Supplementary information Supplementary data are available at Bioinformatics online.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Predicting and explaining the impact of genetic disruptions and interactions on organismal viability

2022

View full text Add to dashboard Cite

show abstract

“…They discuss current approaches -both experimental and computational -that use big data to identify genetic interactions both in humans and model organisms [1]. Detecting sources of bias in transcriptomic data is essential to determine signals of biological significance.…”

Section: Doi: 101515/jib-2017-0052mentioning

confidence: 99%

Biological Big Bytes: Integrative Analysis of Large Biological Datasets

Chen

Harrison

Shanahan

et al. 2017

Journal of Integrative Bioinformatics

View full text Add to dashboard Cite

This current special issue aims to provide a forum for researchers to present the latest advances and state-of-theart techniques, tools and applications in biological big data analysis. The past decade has witnessed tremendous growth in massive and complex omic data sets that are generated continuously by high-throughput experimental technologies such as Next Generation Sequencing (NGS). Today, we are surrounded by a world of big data and we are struggling to make sense of it. The increasing availability of omics data represents an unprecedented opportunity for bioinformaticians, but also a major challenge because of diverse heterogeneous factors. Uncovering hidden patterns from such a huge and heterogeneous amount of omics data allows the creation of predictive models for real-life applications. Research in bioinformatics is moving from a hypothesis-driven to a data-driven approach. The main issue is the need of improved bioinformatics solutions. For example, how to build an appropriate architecture for specific omics data system to manage, access and interpret data. Therefore, new paradigms are needed to deal with omics data, for its annotation and integration and finally for inferring knowledge, answering biological questions and making it available to biologists.For this special issue, we selected one review and six original articles: Benstead-Hume et al. describe how genetic interactions are being therapeutically exploited to identify novel targeted treatments for cancer. They discuss current approaches -both experimental and computational -that use big data to identify genetic interactions both in humans and model organisms [1]. Detecting sources of bias in transcriptomic data is essential to determine signals of biological significance. Alnasir and Shanahan [2] outline a novel method to detect sequence specific bias in short read NGS data, based on determining intra-exon correlations between specific motifs using the big data analysis platform Spark. A Miniature Inverted-repeat Transposable Element (MITE) is a short transposable element, carrying no protein-coding regions. Ge et al. [3] developed MUSTv2 to represent an accurate detection program of recently active MITE copies. Orlov et al. [4] present an analysis of alternative splicing events on an example of glioblastoma cell culture samples using a set of computer tools and integration of the databases. Based on RNA-seq analysis, Babenko et al. revealed a highly statistically significant level increase of ApoE expression in the hypothalamus in chronically aggressive and defeated mice compared to the control. Correlation analysis revealed a close association of ApoE expression profile and proopiomelanocortin gene in the hypothalamus, implying the putative neuroendocrine stress response background of ApoE expression elevation therein [5]. Wang et al. present a generalized approach for measuring relationships between any pairs of genes, which is based on statistical prediction. They derived two particular versions of the generalized approach, least square...

show abstract

“…Unfortunately, genetic interactions are not highly conserved between lower eukaryotes and their human orthologue equivalents [26]. Instead, in order to identify novel human SSL interactions, we are left to infer and predict these pairs indirectly from existing human and model organism data through the use of models and other computational techniques [27].…”

Section: Introductionmentioning

confidence: 99%

Predicting synthetic lethal interactions using conserved patterns in protein interaction networks

Benstead-Hume¹,

Chen²,

Hopkins³

et al. 2019

PLoS Comput Biol

Self Cite

View full text Add to dashboard Cite

In response to a need for improved treatments, a number of promising novel targeted cancer therapies are being developed that exploit human synthetic lethal interactions. This is facilitating personalised medicine strategies in cancers where specific tumour suppressors have become inactivated. Mainly due to the constraints of the experimental procedures, relatively few human synthetic lethal interactions have been identified. Here we describe SLant (Synthetic Lethal analysis via Network topology), a computational systems approach to predicting human synthetic lethal interactions that works by identifying and exploiting conserved patterns in protein interaction network topology both within and across species. SLant out-performs previous attempts to classify human SSL interactions and experimental validation of the models predictions suggests it may provide useful guidance for future SSL screenings and ultimately aid targeted cancer therapy development.

show abstract

Computational Approaches to Identify Genetic Interactions for Cancer Therapeutics

Cited by 5 publications

References 97 publications

Predicting and explaining the impact of genetic disruptions and interactions on organismal viability

Predicting and explaining the impact of genetic disruptions and interactions on organismal viability

Biological Big Bytes: Integrative Analysis of Large Biological Datasets

Predicting synthetic lethal interactions using conserved patterns in protein interaction networks

Contact Info

Product

Resources

About