LGscore: A method to identify disease-related genes using biological literature and Google data

Abstract

Since the genome project in 1990s, a number of studies associated with genes have been conducted and researchers have confirmed that genes are involved in disease. For this reason, the identification of the relationships between diseases and genes is important in biology. We propose a method called LGscore, which identifies disease-related genes using Google data and literature data. To implement this method, first, we construct a disease-related gene network using text-mining results. We then extract gene-gene interactions based on co-occurrences in abstract data obtained from PubMed, and calculate the weights of edges in the gene network by means of Z-scoring. The weights contain two values: the frequency and the Google search results. The frequency value is extracted from literature data, and the Google search result is obtained using Google. We assign a score to each gene through a network analysis. We assume that genes with a large number of links and numerous Google search results and frequency values are more likely to be involved in disease. For validation, we investigated the top 20 inferred genes for five different diseases using answer sets. The answer sets comprised six databases that contain information on disease-gene relationships. We identified a significant number of disease-related genes as well as candidate genes for Alzheimer's disease, diabetes, colon cancer, lung cancer, and prostate cancer. Our method was up to 40% more accurate than existing methods.

OCDB: a database collecting genes, miRNAs and drugs for obsessive-compulsive disorder

“…We defined a score method to measure the relevance (specificity) of the information in our system to OCD. The concept of relevance of an item A (gene, miRNA, drug, SNP, region) to the pathology OCD is defined through a triplet of values normalized by z-scores ( 63 ): w 1 : the number of articles in OCDB containing A and the association with the pathology OCD. w 2 : the number of relations (A, B) in the database.…”

Expanding a Database-derived Biomedical Knowledge Graph via Multi-relation Extraction from Biomedical Abstracts

“…For example, a possible extractor would say gene X is associated with disease Y, because gene X and disease Y appear together more often than individually [20]. This approach has been used to establish the following relationship types: disease-gene relationships [20,21,22,23,24,25], protein-protein interactions [24,26,27], drug-disease treatments [28], and tissue-gene relations [29]. Extractors using the co-occurrence strategy provide exceptional recall results; however, these methods may fail to detect underreported relationships, because they depend on entity-pair frequency for detection.…”

The research on gene-disease association based on text-mining of PubMed

“…Text mining method had been applied in studying various biological problems like functional genomics [ 9 ], biological pathways [ 10 ], protein-protein interactions [ 11 ], protein representation [ 12 ], drug-gene association [ 13 ], comparative toxicogenomics [ 14 , 15 ], neuropsychiatric disorder [ 16 ], and other areas in the biomedical domain [ 17 ] including large-scale bioinformatics analyses [ 8 , 18 – 32 ]. DISEASES predicted the association through the co-occurrence method [ 21 ].…”

“…MimMiner [ 28 ] transformed OMIM [ 29 ] text to a relationship matrix and quantified the association among diseases using the term frequency–inverse document frequency method (TF-IDF). CATAPULT [ 8 ] and Heterogeneous Network Edge Prediction (HNEP) [ 30 ] integrated the graphic model and machine learning method, IMC [ 31 ] used a semi-supervised machine learning method, and LGscore [ 32 ] associated genes with disease through a Google search engine to predict associations between genes and diseases.…”