Related Experiment Video

Updated: Nov 23, 2025

Navigating MARRVEL, a Web-Based Tool that Integrates Human Genomics and Model Organism Genetics Information

Navigating MARRVEL, a Web-Based Tool that Integrates Human Genomics and Model Organism Genetics Information

Published on: August 15, 2019

svMIL: predicting the pathogenic effect of TAD boundary-disrupting somatic structural variants through multiple

Marleen M Nieboer¹, Jeroen de Ridder¹

¹Center for Molecular Medicine, Oncode Institute, University Medical Center Utrecht, Utrecht 3584 CG, The Netherlands.

Bioinformatics (Oxford, England)

|December 31, 2020

Abstract

Motivation:

Despite the fact that structural variants (SVs) play an important role in cancer, methods to predict their effect, especially for SVs in non-coding regions, are lacking, leaving them often overlooked in the clinic. Non-coding SVs may disrupt the boundaries of Topologically Associated Domains (TADs), thereby affecting interactions between genes and regulatory elements such as enhancers. However, it is not known when such alterations are pathogenic. Although machine learning techniques are a promising solution to answer this question, representing the large number of interactions that an SV can disrupt in a single feature matrix is not trivial.

Results:

We introduce svMIL: a method to predict pathogenic TAD boundary-disrupting SV effects based on multiple instance learning, which circumvents the need for a traditional feature matrix by grouping SVs into bags that can contain any number of disruptions. We demonstrate that svMIL can predict SV pathogenicity, measured through same-sample gene expression aberration, for various cancer types. In addition, our approach reveals that somatic pathogenic SVs alter different regulatory interactions than somatic non-pathogenic SVs and germline SVs.

Availability And Implementation:

All code for svMIL is publicly available on GitHub: https://github.com/UMCUGenetics/svMIL.

Supplementary Information:

Supplementary data are available at Bioinformatics online.

Insights

Predicting the impact of structural variants (SVs) in cancer is challenging. Our new method, svMIL, uses multiple instance learning to identify pathogenic SVs disrupting Topologically Associated Domains (TADs) and their regulatory interactions.

Area of Science:

Genomics
Cancer Biology
Bioinformatics

Background:

Structural variants (SVs) are crucial in cancer development, but predicting their functional impact, particularly in non-coding regions, remains a significant challenge.
Non-coding SVs can disrupt Topologically Associated Domain (TAD) boundaries, altering gene-regulatory element interactions, yet their pathogenicity is often unclear.
Current machine learning approaches struggle to represent the complex interaction disruptions caused by SVs within traditional feature matrices.

Purpose of the Study:

To develop a novel method, svMIL, for predicting the pathogenicity of SVs that disrupt TAD boundaries.
To overcome the limitations of traditional feature matrices in representing SV-induced interaction disruptions using multiple instance learning.
To analyze the distinct regulatory interaction alterations associated with pathogenic somatic SVs compared to non-pathogenic somatic and germline SVs.

Main Methods:

Introduction of svMIL, a multiple instance learning (MIL) based method for predicting pathogenic SV effects.
svMIL groups SVs into 'bags,' accommodating varying numbers of disrupted interactions without requiring a fixed feature matrix.
Validation of svMIL's predictive capability using gene expression aberration data from various cancer types.

Main Results:

svMIL successfully predicts the pathogenicity of TAD boundary-disrupting SVs across different cancer types, correlating with gene expression aberrations.
The study reveals distinct patterns of regulatory interaction disruption between pathogenic somatic SVs, non-pathogenic somatic SVs, and germline SVs.
svMIL demonstrates the utility of MIL in handling complex genomic interaction data for pathogenicity prediction.

Conclusions:

svMIL provides a robust method for identifying pathogenic SVs in non-coding regions by analyzing their impact on TAD boundaries and regulatory interactions.
The findings highlight the differential impact of various SV types on regulatory networks, offering insights into cancer mechanisms.
The developed method and its code are publicly available, facilitating further research in cancer genomics and SV analysis.

Read full analysis

More Related Videos

In Vivo Functional Study of Disease-associated Rare Human Variants Using Drosophila

In Vivo Functional Study of Disease-associated Rare Human Variants Using Drosophila

Published on: August 20, 2019

In Vivo Modeling of the Morbid Human Genome using Danio rerio

In Vivo Modeling of the Morbid Human Genome using Danio rerio

Published on: August 24, 2013

Related Concept Videos

Pleiotropy

Pleiotropy

Pleiotropy is the phenomenon in which a single gene impacts multiple, seemingly unrelated phenotypic traits. For example, defects in the SOX10 gene cause Waardenburg Syndrome Type 4, or WS4, which can cause defects in pigmentation, hearing impairments, and an absence of intestinal contractions necessary for elimination. This diversity of phenotypes results from the expression pattern of SOX10 in early embryonic and fetal development. SOX10 is found in neural crest cells that form melanocytes,...

Multiple Allele Traits

Multiple Allele Traits

The Concept of Multiple Allelism

Genome-wide Association Studies-GWAS

Genome-wide Association Studies-GWAS

Genome-wide association studies or GWAS are used to identify whether common SNPs are associated with certain diseases. Suppose specific SNPs are more frequently observed in individuals with a particular disease than those without the disease. In that case, those SNPs are said to be associated with the disease. Chi-square analysis is performed to check the probability of the allele likely to be associated with the disease.
GWAS does not require the identification of the target gene involved in...

Comparing Copy Number Variations and SNPs

Comparing Copy Number Variations and SNPs

Sequencing of the human genome has opened up several best-kept secrets of the genome. Scientists have identified thousands of genome variations that exist within a population. These variations can be a single nucleotide or a larger chromosomal variation.
Copy number variations or CNVs are the structural variations that cover more than 1kb of DNA sequence. The single nucleotide polymorphism (SNP), on the other hand, is a single nucleotide change or a point mutation that is found in more than 1%...

Cancers Originate from Somatic Mutations in a Single Cell

Cancers Originate from Somatic Mutations in a Single Cell

Cancer arises from mutations in the critical genes that allow healthy cells to escape cell cycle regulation and acquire the ability to proliferate indefinitely. Though originating from a single mutation event in one of the originator cells, cancer progresses when the mutant cell lines continue to gain more and more mutations, and finally, become malignant. For example, chronic myelogenous leukemia (CML) develops initially as a non-lethal increase in white blood cells, which progressively...

Epistasis Analysis

Epistasis Analysis

Although Mendel chose seven unrelated traits in peas to study gene segregation, most traits involve multiple gene interactions that create a spectrum of phenotypes. When the interaction of various genes or alleles at different locations influences a phenotype, this is called epistasis. Epistasis often involves one gene masking or interfering with the expression of another (antagonistic epistasis). Epistasis often occurs when different genes are part of the same biochemical pathway. The...

Articles linked to this work by shared authors, journal, and citation graph.

Large-scale simulation of coverage and error rate tradeoffs for cancer detection in cell-free DNA whole-genome sequencing.

Bioinformatics (Oxford, England)·2026

How 'Sturgeon' guides the surgeon in pediatric neuro-oncology.

Neuro-oncology·2026

Univariate- and machine learning-based plasma metabolite signature differentiates PSC-IBD from IBD and is predicted to be driven by gut microbial changes.

Metabolomics : Official journal of the Metabolomic Society·2026

Regulatory grammar in human promoters uncovered by MPRA-based deep learning.

Nature·2026

Bacterial cell-free DNA profiling reveals the co-elevation of multiple bacteria in newborn foals with suspected sepsis.

iScience·2026

A computational framework to dissect imputation strategies for single-cell histone modification data.

NAR genomics and bioinformatics·2025

See all related articles