Mining disease-specific molecular association profiles from biomedical literature

Li, Jiao; Zhu, Xiaoyan; Chen, Jake Y.

doi:10.1145/1363686.1363984

Cited by 5 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From the retrieved abstract collection T NET , the drugs { d 1 , d 2 ,…, d n } were then identified automatically by our system combining both dictionary and rule directives. We calculated a p-value for each drug d j in T NET using methods described in [53] and later derived its false discovery rate. Let the null hypothesis H 0 be that document frequency of drug d j in T NET come from a random distribution.…”

Section: Methodsmentioning

confidence: 99%

“…In this work, we first retrieved the entire PubMed abstracts using the expanded list { p 1 , p 2 ,…, p m } containing all the proteins in a network as the initial query [53] . From the retrieved abstract collection T NET , the drugs { d 1 , d 2 ,…, d n } were then identified automatically by our system combining both dictionary and rule directives.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Building Disease-Specific Drug-Protein Connectivity Maps from Molecular Interaction Networks and PubMed Abstracts

2009

Self Cite

View full text Add to dashboard Cite

The recently proposed concept of molecular connectivity maps enables researchers to integrate experimental measurements of genes, proteins, metabolites, and drug compounds under similar biological conditions. The study of these maps provides opportunities for future toxicogenomics and drug discovery applications. We developed a computational framework to build disease-specific drug-protein connectivity maps. We integrated gene/protein and drug connectivity information based on protein interaction networks and literature mining, without requiring gene expression profile information derived from drug perturbation experiments on disease samples. We described the development and application of this computational framework using Alzheimer's Disease (AD) as a primary example in three steps. First, molecular interaction networks were incorporated to reduce bias and improve relevance of AD seed proteins. Second, PubMed abstracts were used to retrieve enriched drug terms that are indirectly associated with AD through molecular mechanistic studies. Third and lastly, a comprehensive AD connectivity map was created by relating enriched drugs and related proteins in literature. We showed that this molecular connectivity map development approach outperformed both curated drug target databases and conventional information retrieval systems. Our initial explorations of the AD connectivity map yielded a new hypothesis that diltiazem and quinidine may be investigated as candidate drugs for AD treatment. Molecular connectivity maps derived computationally can help study molecular signature differences between different classes of drugs in specific disease contexts. To achieve overall good data coverage and quality, a series of statistical methods have been developed to overcome high levels of data noise in biological networks and literature mining results. Further development of computational molecular connectivity maps to cover major disease areas will likely set up a new model for drug development, in which therapeutic/toxicological profiles of candidate drugs can be checked computationally before costly clinical trials begin.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Building Disease-Specific Drug-Protein Connectivity Maps from Molecular Interaction Networks and PubMed Abstracts

2009

Self Cite

View full text Add to dashboard Cite

show abstract

“…To construct protein-drug association network in cancer context, we applied a biomedical literature mining method based on a recent approach for building disease-specific molecular association profiles [8]. Protein-drug associations for cancer c 1 and c 2 , PDA 1 and PDA 2 , were constructed independently following the same workflow.…”

Section: Pda Constructionmentioning

confidence: 99%

“…First, we extended the work in [8], to expand and rank-order proteins in the context of molecular interaction sub-networks for cancers built on the work of [9], before applying text mining techniques to identify novel associations between cancer-relevant genes/proteins and anticancer drugs. The integration of biomedical literature mining and network data analysis enables us to generate protein-drug association profiles of both statistical and biological significance computationally.…”

Section: Introductionmentioning

confidence: 99%

Construct anticancer drug-drug correlation network

Crowell

Chen

2009

Proceedings of the 2009 ACM Symposium on Applied Computing

Self Cite

View full text Add to dashboard Cite

Network biology methods have been promising in linking diseases, genes, and drugs. In this study, we propose a novel computational method to construct drug-drug correlation networks, which consist of drug compounds as network graph nodes and correlated proteindrug profiles above a pre-determined threshold as network graph edges. This computational method is based on extensions of related work on identifying disease-specific proteins within protein-protein sub-networks, and mining protein-drug association profiles from biomedical literature. Our method provides a quantitative framework to compare how each drug compound from one therapeutic area is correlated with other drug compounds in other therapeutic areas, using a drug's protein-drug association profile mined from biomedical literature. We applied this method to the study of drug re-purposing and found that two breast cancer drugs "Mitomycin" and "Bleomycin" may be top drug candidates for treating pancreatic cancers.

show abstract

“…The copyright holder for this preprint this version posted February 10, 2023. ; https://doi.org/10.1101/2023.02.09.527562 doi: bioRxiv preprint 3 been used to curate significant genes related to a biological condition [1,11,12]. However, results may be biased toward literature studies that are "popular" and not necessarily directly indicative of biological importance and significance [1,8].…”

Section: Introductionmentioning

confidence: 99%

Prioritizing Complex Disease Genes from Heterogeneous Public Databases

Gong

Chen

2023

Preprint

View full text Add to dashboard Cite

Background: Complex human diseases are defined not only by sophisticated patterns of genetic variants/mutations upstream but also by many interplaying genes, RNAs, and proteins downstream. Analyzing multiple genomic and functional genomic data types to determine a short list of genes or molecules of interest is a common task called ″gene prioritization″ in biology. There are many statistical, biological, and bioinformatic methods developed to perform gene prioritization tasks. However, little research has been conducted to examine the relationships among the technique used, merged/separate use of each data modality, the gene list′s network/pathway context, and various gene ranking/expansions. Methods: We introduce a new analytical framework called ″Gene Ranking and Iterative Prioritization based on Pathways″ (GRIPP) to prioritize genes derived from different modalities. Multiple data sources, such as CBioPortal, PAGER, and COSMIC were used to compile the initial gene list. We used the PAGER software to expand the gene list based on biological pathways and the BEERE software to construct protein-protein interaction networks that include the gene list to rank order genes. We produced a final gene list for each data modality iteratively from an initial draft gene list, using glioblastoma multiform (GBM) as a case study. Conclusion: We demonstrated that GBM gene lists obtained from three modalities (differential gene expressions, gene mutations, and copy number alterations) and several data sources could be iteratively expanded and ranked using GRIPP. While integrating various modalities of data can be useful to generate an integrated ranked gene list related to any specific disease, the integration may also decrease the overall significance of ranked genes derived from specific data modalities. Therefore, we recommend carefully sorting and integrating gene lists according to each modality, such as gene mutations, epigenetic controls, or differential expressions, to procure modality-specific biological insights into the prioritized genes.

show abstract

Mining disease-specific molecular association profiles from biomedical literature

Cited by 5 publications

References 17 publications

Building Disease-Specific Drug-Protein Connectivity Maps from Molecular Interaction Networks and PubMed Abstracts

Building Disease-Specific Drug-Protein Connectivity Maps from Molecular Interaction Networks and PubMed Abstracts

Construct anticancer drug-drug correlation network

Prioritizing Complex Disease Genes from Heterogeneous Public Databases

Contact Info

Product

Resources

About