SLGNN: synthetic lethality prediction in human cancers based on factor-aware knowledge graph neural network

Abstract

Motivation Synthetic Lethality (SL) is a form of genetic interaction that can selectively kill cancer cells without damaging normal cells. Exploiting this mechanism is gaining popularity in the field of targeted cancer therapy and anticancer drug development. Due to the limitations of identifying SL interactions from laboratory experiments, an increasing number of research groups are devising computational prediction methods to guide the discovery of potential SL pairs. Although existing methods have attempted to capture the underlying mechanisms of SL interactions, methods that have a deeper understanding of and attempt to explain SL mechanisms still need to be developed. Results In this work, we propose a novel SL prediction method, SLGNN. This method is based on the following assumption: SL interactions are caused by different molecular events or biological processes, which we define as SL-related factors that lead to SL interactions. SLGNN, apart from identifying SL interaction pairs, also models the preferences of genes for different SL-related factors, making the results more interpretable for biologists and clinicians. SLGNN consists of three steps: first, we model the combinations of relationships in the gene-related knowledge graph as the SL-related factors. Next, we derive initial embeddings of genes through an explicit message aggregation process of the knowledge graph. Finally, we derive the final gene embeddings through an SL graph, constructed using known SL gene pairs, utilizing factor-based message aggregation. At this stage, a supervised end-to-end training model is used for SL interaction prediction. Based on experimental results, the proposed SLGNN model outperforms all current state-of-the-art SL prediction methods and provides better interpretability. Availability SLGNN is freely available at https://github.com/zy972014452/SLGNN. Supplementary information Supplementary data are available at Bioinformatics online.

MPASL: multi-perspective learning knowledge graph attention network for synthetic lethality prediction in human cancer

“… (5) KG4SL ( Wang et al, 2021 ) represents the first novel SL interaction prediction model based on knowledge graphs and graph neural networks, effectively leveraging rich semantic information encoded in KGs. (6) SLGNN ( Zhu et al, 2023 ) is a factor-aware knowledge graph neural network for learning gene embeddings and predicting SL interactions. …”

“…To overcome these limitations, a several computational methods have been developed for SL prediction. These methods fall into two categories: (i) knowledge-based methods and (ii) supervised machine-learning methods ( Zhu et al, 2023 ). Knowledge-based methods rely on prior knowledge or assumptions (i.e., gene mutations ( Lu et al, 2020 ) or CNVs ( Lu et al, 2018 )) to detect SL pairs.…”

“…It uses independent knowledge embeddings to capture the underlying biological mechanisms of interconnected SL pairs. Zhu et al ( Zhu et al, 2023 ) utilized relations in knowledge graphs to represent SL-related factors and learned latent representations of genes through message aggregation. It is evident that employing KG entities such as gene, pathway and their neighbors yields a more accurate embedding representation, but previously KG-based methods ignore the preferences of existing SL interactions and layer-wise differences of entities.…”

Petagraph: A large-scale unifying knowledge graph framework for integrating biomolecular and biomedical data

“…Knowledge graphs have been used for many applications, including drug discovery and repurposing, target detection, and prediction (Alshahrani and Hoehndorf 2018;Moon et al 2021;Zheng et al 2021;Alves et al 2021). Other applications include integration and analysis of heterogeneous COVID-19 data (Steenwinckel et al 2020;Cernile et al 2021;Reese et al 2021;Domingo-Fernández et al 2021;Zhang et al 2021;Ostaszewski et al 2021;Chen et al 2021), oncology research (Zhu et al 2023;Zhao et al 2023;Jha et al), and gene-disease associations (Choi and Lee 2021;Alves et al 2021) among many others. The unifying intent in these projects is the meaningful The data loading pipeline contains four major steps (Figure 2): (1) Data selection and modeling, (2) Cleaning and preparing data, (3) Executing the UBKG "OWLNETS" pythons scripts and (4) Import using the Neo4j Bulk Import tool.…”

Cell context-specific Synthetic lethality Prediction and Mechanism Analysis

“…Firstly, most models integrate the SL gene pairs from the population level as positive labels [35][36][37][38][39][40][41][42][43][44][45][46][47][48]. Population-level data is derived from a set of cell lines or a collection of tumor samples obtained from multiple patients [49].…”