Biomedical Hypothesis Generation by Text Mining and Gene Prioritization

Abstract

Text mining methods can facilitate the generation of biomedical hypotheses by suggesting novel associations between diseases and genes. Previously, we developed a rare-term model called RaJoLink (Petric et al, J. Biomed. Inform. 42(2): 219-227, 2009) in which hypotheses are formulated on the basis of terms rarely associated with a target domain. Since many current medical hypotheses are formulated in terms of molecular entities and molecular mechanisms, here we extend the methodology to proteins and genes, using a standardized vocabulary as well as a gene/protein network model. The proposed enhanced RaJoLink rare-term model combines text mining and gene prioritization approaches. Its utility is illustrated by finding known as well as potential gene-disease associations in ovarian cancer using MEDLINE abstracts and the STRING database.

A systematic review on literature-based discovery workflow

“…Since most of the LBD research performed in medicine, the most common term representation is using UMLS and MeSH (Lever et al, 2017;Preiss & Stevenson, 2017). Apart from these two medical resources, other medical databases such as Entrez Gene , HUGO (Petric et al, 2014), LocusLink (Hristovski et al, 2005), OMIM (Hristovski et al, 2003) and PharmGKB (Kim & Park, 2016) have also being used extract data units. LBD studies in other domains mainly consider word or word phrases (n-grams) as their term representation (Qi & Ohsawa, 2016) that have been extracted using techniques such as Part-Of-Speech (POS) tag patterns.…”

“…Network/Graph-based Measures: Network/graph-based measures analyse node and edgelevel attributes to score the associations. Examples of measures that represent this category include Degree centrality (Goodwin, Cohen & Rindflesch, 2012), Eigenvector centrality (Özgür et al, 2010), Closeness centrality (Özgür et al, 2011), Betweenness centrality (Özgür et al, 2010), Common Neighbours (Kastrin, Rindflesch & Hristovski, 2014), Jaccard Index (Kastrin, Rindflesch & Hristovski, 2014), Preferential Attachment (Kastrin, Rindflesch & Hristovski, 2014), Personalised PageRank (Petric et al, 2014), Personalised Diffusion Ranking (Petric et al, 2014), and Spreading Activation (Goodwin, Cohen & Rindflesch, 2012). Knowledge-based Measures: This category denotes the scoring measures such as MeSHbased Literature cohesiveness (Swanson, Smalheiser & Torvik, 2006), semantic type cooccurrence (Jha & Jin, 2016b), chemDB atomic count (Ijaz, Song & Lee, 2010), and chemDB XLogP (Ijaz, Song & Lee, 2010) that involve the knowledge from structured resources to rank the associations.…”

“…Resources to acquire background domain knowledge: The main purposes of extracting the domain knowledge are; (1) input data preparation (e.g., concept extraction), (2) filtering the noisy, uninteresting or unrelated associations (e.g., semantic type filtering), (3) prepare a ranking mechanism (e.g., hierarchical ranking), (4) evaluate the results (e.g., compare results with curated databases), and (5) training data preparation. The popular domain dependent resources used in the discipline are; -UMLS: (Lever et al, 2017;Vlietstra et al, 2017;Preiss & Stevenson, 2017) -MeSH: (Baek et al, 2017;Xun et al, 2017;Pusala et al, 2017) -SemMedDB/Semantic Medline: (Vlietstra et al, 2017;Cairelli et al, 2015) -Gene Ontology: (Baek et al, 2017;Huang et al, 2016; -Entrez Gene Database: (Baek et al, 2017;Liang, Wang & Wang, 2013;Kwofie et al, 2011) -Kyoto Encyclopedia of Genes and Genomes (KEGG): (Kwofie et al, 2011) -HGNC/HUGO: (Petric et al, 2014;Ding et al, 2013;Maciel et al, 2011) -UNIPROT: (Baek et al, 2017;Vlietstra et al, 2017), Swiss-Prot (Jelier et al, 2008) -Therapeutic Target Database (TTD): (Yang et al, 2017;Maciel et al, 2011) -LocusLink: (Smalheiser, 2005;Hristovski et al, 2003) -Online Mendelian Inheritance in Man (OMIM) (Hristovski et al, 2003;Wren et al, 2004) -Drug Bank: (Vlietstra et al, 2017;Maciel et al, 2011;Ding et al, 2013) -Comparative Toxicogenomics Database (CTD): (Vlietstra et al, 2017;Yang et al, 2017) -BioGRID: (Huang et al, 2016;Crichton et al, 2018) -Gene2pubmed: (Cheung,...…”

In Silico Exploration of the Potential Role of Acetaminophen and Pesticides in the Etiology of Autism Spectrum Disorder

“…To explore whether APAP and pesticide exposure might influence ASD risk through similar underlying biological mechanisms, we conducted this study to perform database mining of the genetic associations of ASD in combination with network and functional analysis of candidate genes. Computer-based text mining methods have been used before to generate biomedical hypotheses on the impact of multiple exposures by examining novel associations between genes and diseases [ 34 ]. For example, literature mining and computational systems biology methods were used to explore the possible etiologic links between environmental chemicals, genes of interest, and type II diabetes [ 35 ].…”

Peer Review #1 of "A systematic review on literature-based discovery workflow (v0.1)"

“…Network/Graph-based Measures: Network/graph-based measures analyse node and edge-level attributes to score the associations. Examples of measures that represent this category include Degree centrality(Goodwin et al, 2012), Eigenvector centrality(Özgür et al, 2010), Closeness centrality(Özgür et al, 2011), Betweenness centrality(Özgür et al, 2010), Common Neighbours(Kastrin et al, 2014), Jaccard Index(Kastrin et al, 2014), Preferential Attachment(Kastrin et al, 2014), Personalised PageRank(Petric et al, 2014), Personalised Diffusion Ranking(Petric et al, 2014), and Spreading Activation(Goodwin et al, 2012).Knowledge-based Measures: This category denotes the scoring measures such as MeSH-based Literature cohesiveness(Swanson et al, 2006), semantic type co-occurrence(Jha and Jin, 2016b), chemDB atomic count(Ijaz et al, 2010), and chemDB XLogP(Ijaz et al, 2010) that involve the knowledge from structured resources to rank the associations. The advantage of these measures is that they entangle semantic aspects into consideration to decide the potentiality of the association.Relations-based Measures: Relations/predicate based measures (a sub-class of knowledge-based measures) analyse the relations extracted from resources such as SemRep to rank/threshold associations.…”