Automatic extraction of gene-disease associations from literature using joint ensemble learning

Abstract

A wealth of knowledge concerning relations between genes and its associated diseases is present in biomedical literature. Mining these biological associations from literature can provide immense support to research ranging from drug-targetable pathways to biomarker discovery. However, time and cost of manual curation heavily slows it down. In this current scenario one of the crucial technologies is biomedical text mining, and relation extraction shows the promising result to explore the research of genes associated with diseases. By developing automatic extraction of gene-disease associations from the literature using joint ensemble learning we addressed this problem from a text mining perspective. In the proposed work, we employ a supervised machine learning approach in which a rich feature set covering conceptual, syntax and semantic properties jointly learned with word embedding are trained using ensemble support vector machine for extracting gene-disease relations from four gold standard corpora. Upon evaluating the machine learning approach shows promised results of 85.34%, 83.93%,87.39% and 85.57% of F-measure on EUADR, GAD, CoMAGC and PolySearch corpora respectively. We strongly believe that the presented novel approach combining rich syntax and semantic feature set with domain-specific word embedding through ensemble support vector machines evaluated on four gold standard corpora can act as a new baseline for future works in gene-disease relation extraction from literature.

Fig 1. Schematic architecture of the gene-disease relation extraction system.

Fig 2. Extraction workflow of the supervised machine learning approach.

Fig 3. Feature representation of gene-disease relation extraction.

a) The sentence is tagged with both LOXL1 gene and Exfoliation glaucoma disease from EU-ADR corpus with PMCID: PMC2605423 b) Word window representation of syntax and semantic features c)Tokens positioned at the left and right (n-gram) of the candidates(LOXL1 and exfoliation glaucoma)d)Locating the words between the entities for relational and trigger words e) Phrasal feature from the relational word f) Finding context specific word using trigger word templates.

Fig 4. Representation of the skip-gram (SG) model with target word gene at the input layer and the learned contextual words like a promoter, susceptibility, protein etc. are in the output layer, adapted from [48].

Fig 5. ROC with respect to FPR and TPR on four corpora upon 10-fold cross-validation.

In this figure, a, b, c, and d represents the receiver operating curves of EU-ADR, GAD, CoMAGC and PolySearch corpora respectively.

Fig 6. Performance evaluation of gene disease relation extraction on four different corpora.

Fig 7. Performance comparison of gene disease relation extraction on four different corpora.