The social life of bioinformatics tools, an OpenEBench's graph-based network

This project presents a graph-based network of the co-usage, understood as being cited by the same scientific publication, of research software from the collection hosted by OpenEBench, and publications from the bioinformatics' domain. The thesis aims to reflect how the bioinformatics' com...

Full description

Bibliographic Details
Author: Aguiló Castillo, Sergi
Format: master thesis
Publication Date:2021
Country:España
Institution:Universitat Politècnica de Catalunya (UPC)
Repository:UPCommons. Portal del coneixement obert de la UPC
Language:English
OAI Identifier:oai:upcommons.upc.edu:2117/356877
Online Access:https://hdl.handle.net/2117/356877
Access Level:Open access
Keyword:Databases
Bioinformatics
Base de dades
Base de dades orientada a grafs
Neo4j
Anàlisi de grafs
Base de dades relacional
Eines bioinformàtiques
Bioinformàtica
Xarxa social
Benchmarking
Flux de treball
Database
Graph database
Analysis of graphs
Relational database
Bioinformatics' tools
Social network
Workflows
Bases de dades
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica::Bioinformàtica
Description
Summary:This project presents a graph-based network of the co-usage, understood as being cited by the same scientific publication, of research software from the collection hosted by OpenEBench, and publications from the bioinformatics' domain. The thesis aims to reflect how the bioinformatics' community behave with the tools that they create, identify specific communities in the network and provide guidance on scientific tools worth benchmarking in the context of community-led scientific benchmarking efforts. Also, assess if the noise of the non-significant references relationships in a publication can change the behaviour of the network. To do so, we retrieve more than one million scientific publications related to a subset of bioinformatics tools. Then, we infer tools from the publications with the OpenEBench's platform, enriching them with EDAM ontology terms. After, the relationships among tools and publications are created and upload in a graph database. Finally, a clustering approach, a centrality method and a web page for query the relationships of the tools or topics are applied. To assess the noise, we create a graph for the OpenAccess publications and store the relationships by section (Introduction, Methods, Results, and Discussion). The resulting network is able to show the relationships of any tool or publication and topic of the bioinformatics domain that exists in the graph database. Also, it classifies the network in different communities. The graph of the OpenAccess research papers shows that the noise is not significant enough to change the network's connections and that the relationships retrieved are almost equal to the ones from the Methods section.