Related Experiment Video
Updated: Aug 28, 2025
A Data Integration Workflow to Identify Drug Combinations Targeting Synthetic Lethal Interactions
Published on: May 27, 2021
NSF4SL: negative-sample-free contrastive learning for ranking synthetic lethal partner genes in human cancers
Shike Wang1, Yimiao Feng1, Xin Liu1
1School of Information Science and Technology, ShanghaiTech University, Shanghai 201210, China.
Motivation:
Detecting synthetic lethality (SL) is a promising strategy for identifying anti-cancer drug targets. Targeting SL partners of a primary gene mutated in cancer is selectively lethal to cancer cells. Due to high cost of wet-lab experiments and availability of gold standard SL data, supervised machine learning for SL prediction has been popular. However, most of the methods are based on binary classification and thus limited by the lack of reliable negative data. Contrastive learning can train models without any negative sample and is thus promising for finding novel SLs.
Results:
We propose NSF4SL, a negative-sample-free SL prediction model based on a contrastive learning framework. It captures the characteristics of positive SL samples by using two branches of neural networks that interact with each other to learn SL-related gene representations. Moreover, a feature-wise data augmentation strategy is used to mitigate the sparsity of SL data. NSF4SL significantly outperforms all baselines which require negative samples, even in challenging experimental settings. To the best of our knowledge, this is the first time that SL prediction is formulated as a gene ranking problem, which is more practical than the current formulation as binary classification. NSF4SL is the first contrastive learning method for SL prediction and its success points to a new direction of machine-learning methods for identifying novel SLs.
Availability And Implementation:
Our source code is available at https://github.com/JieZheng-ShanghaiTech/NSF4SL.
Supplementary Information:
Supplementary data are available at Bioinformatics online.
Insights
NSF4SL, a novel machine learning model, predicts synthetic lethality (SL) without negative samples. This contrastive learning approach outperforms existing methods, offering a new direction for identifying anti-cancer drug targets.
Area of Science:
- Computational Biology and Bioinformatics
- Genomics and Precision Medicine
- Machine Learning in Drug Discovery
Background:
- Synthetic lethality (SL) is a key strategy for discovering anti-cancer drug targets by exploiting genetic vulnerabilities in cancer cells.
- Current supervised machine learning models for SL prediction are often limited by the scarcity of reliable negative data and the binary classification approach.
- Contrastive learning offers a promising alternative by enabling model training without negative samples, facilitating the identification of novel SLs.
Purpose of the Study:
- To develop a novel, negative-sample-free model for synthetic lethality (SL) prediction using a contrastive learning framework.
- To address the limitations of existing binary classification methods by formulating SL prediction as a gene ranking problem.
- To improve the identification of potential anti-cancer drug targets by leveraging machine learning for SL detection.
Main Methods:
- Proposed NSF4SL, a negative-sample-free SL prediction model employing a contrastive learning framework.
- Utilized a two-branch neural network architecture where branches interact to learn gene representations relevant to SL.
- Implemented a feature-wise data augmentation strategy to enhance model performance with sparse SL data.
Main Results:
- NSF4SL significantly outperformed existing baseline methods that rely on negative samples, even under challenging data conditions.
- The model demonstrated the effectiveness of contrastive learning for SL prediction, establishing a new benchmark.
- Formulating SL prediction as a gene ranking problem proved more practical and effective than traditional binary classification.
Conclusions:
- NSF4SL represents a significant advancement in machine learning for synthetic lethality prediction, offering a robust alternative to negative-sample-dependent methods.
- The success of NSF4SL highlights the potential of contrastive learning for discovering novel synthetic lethal interactions and anti-cancer drug targets.
- This work pioneers the application of contrastive learning and gene ranking in the field of synthetic lethality prediction.
Related Concept Videos
01:05Cancer-Critical Genes II: Tumor Suppressor Genes
When the function of certain critical genes, especially those involved in cell cycle regulation and cell growth signaling cascades, gets disrupted, it upsets the cell cycle progression. Such cells with unchecked cell cycles start proliferating uncontrollably and eventually develop into tumors.
Such genes that act...
01:33Cancer-Critical Genes I: Proto-oncogenes
When the function of certain critical genes, especially those involved in cell cycle regulation and cell growth signaling cascades, gets disrupted, it upsets the cell cycle progression. Such cells with unchecked cell cycles start proliferating uncontrollably and eventually develop into tumors.
Such genes that act...
02:41Lethal Alleles
Lucien Cuénot discovered lethal alleles in 1905 while studying the inheritance of coat color in mice. The agouti gene is responsible for the color of the coat in mice. This gene codes for an agouti-signaling protein, which is responsible for melanin distribution in mammals. The wild-type allele gives rise to gray-brown coat color in mice, while the mutant allele gives rise to yellow coat color. In addition to coat color, the agouti gene is associated with the yellow...
01:21Cancer Survival Analysis
02:26Protein Networks
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
02:43Mouse Models of Cancer Study
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...