AarXiv8 Aug 2021Cited 2×

MuCoMiD: A Multitask Convolutional Learning Framework for miRNA-Disease Association Prediction.

Thi Ngan Dong, Megha Khosla

Abstract

—Growing evidence from recent studies implies that microRNA or miRNA could serve as biomarkers in various complex human diseases. Since wet-lab experiments are expensive and time-consuming, computational techniques for miRNA-disease association prediction have attracted a lot of attention in recent years. Data scarcity is one of the major challenges in building reliable machine learning models. Data scarcity combined with the use of precalculated hand-crafted input features has led to problems of overfitting and data leakage.

A figure from MuCoMiD: A Multitask Convolutional Learning Framework for miRNA-Disease Association Prediction. — fig. from the paper

§ The Valyu brief

The paper, in five minutes

Read from the full text via Valyu Contents

MuCoMiD, a multitask graph convolutional framework, boosts miRNA‑disease association prediction by ≥3% on benchmark CV and ≥35% on independent tests.

What’s new

i.Introduced MuCoMiD, a novel multitask graph convolutional network for miRNA‑disease association prediction.
ii.Integrated five heterogeneous biological information sources for automatic feature extraction.
iii.Conducted extensive experiments on standard benchmarks (HMDDv2.0, HMDDv3.0) and larger independent test sets, demonstrating significant performance gains.
iv.Released code publicly for reproducibility and future research.

How they did it

MuCoMiD employs a multi‑task graph convolutional architecture that jointly learns from miRNA‑disease, miRNA‑gene, gene‑gene, miRNA family, and disease ontology networks. The model automatically extracts features from these heterogeneous graphs, avoiding hand‑crafted inputs. Performance was evaluated using 5‑fold cross‑validation on benchmark datasets and on independent test sets containing unseen miRNAs and diseases.

What they found

At least 3% improvement in 5‑fold cross‑validation on HMDDv2.0 and HMDDv3.0 datasets compared to state‑of‑the‑art methods.
At least 35% improvement on larger independent test sets with unseen miRNAs and diseases over existing approaches.

§ The Valyu brief

Reading the full paper and taking notes. This takes a few seconds…

§ Ask this paper

Ask a question about this paper

Valyu reads the full text and answers from what the paper actually says.

§ Related work in other archives

Similar papers in all four archives

The same idea searched in arXiv, PubMed, bioRxiv and chemRxiv at once, so a method can lead you to where it is being applied.

AarXiv3 found

PPubMed4 found

BbioRxiv4 found

CchemRxiv2 found

Searching the other archives…