In the last decades, bioinformatics has become one of the key disciplines related with biology and medicine. This field integrates the study of transcriptomics, genomics, proteomics, pharmacogenomics, medical imaging … Bioinformatics has burst into research centres, hospitals and industry, and it’s here to stay.

 

 

Bioinformatics experts have reunited in Granada (Spain) on 14th-16th November, at XIV Symposium on Bioinformatics (JBI 2018). It’s a biannual symposium to bring up on discussion the latest advances on this field on a global scale. This congress has been co-organized by National Bioinformatics Institute (INB/ELIXIR-ES), University of Granada, Centre for Genomics and Oncological Research (GENyO), Barcelona Supercomputing Center (BSC).

The aim for these conferences devoted to bioinformatics is the presentation and discussion of the latest developments in bioinformatics among its multiple disciplines: structural, functional, translational, systems biology …

 

From pharmacogenomics to trendy geometrical shapes

Julio Sáez-Rodríguez, professor of Medical Bioinformatis and Data Analysis at the Faculty of Medicine of the University of Heidelberg, opened the symposium with a talk about computational personalized medicine in cancer. He focused on tumour alterations that confer them with resistance or sensibility to different treatments. The analysis of this pharmacogenomics data can provide of new biomarkers and offer new combination therapies. Statistics and machine learning can help us, but in most of the cases the predictability is low. This is caused by the lack of samples or the need to analyze other molecular layers, as proteomics. Professor Sáez-Rodríguez has developed a series of software to improve this modelings, such as CellNOpt, PROGENy, DoRothEA or CARNIVAL.

Relevance of regional genomics models

Another speaker following the line of personalized medicine and genomics was José Mg Izarzugaza, associate professor at the Center for Biological Sequence Analysis in Denmark. There he carries out the GenomeDenmark project. It aims to establish a regional reference genome to improve future personalized medicine studies and ease its clinical application. The availability of a regional reference allows an integration of genetic variants specific to a population. This can minimize errors and false discoveries that arise when comparing to a less specific global reference.

Also about personalized medicine with regional perspective works Joaquín Dopazo, head of Clinical Bioinformatics Area from the Fundación Progreso y Salud (FPS). Currently, they are developing a project aimed to ease the use of genomics data for diagnosis and treatment in the national health system. Around 12.000 polimorfisms have been detected in spanish population, not included in external databases. Using this local data increase the filtering of variants efficiency, like within the GenomeDenmark project. The Clinical Bioinformatics Area has also developed the Collaborative Spanish Variant Server (CSVS), a growing platform offering accurate information about variants specific of spanish population.

Bioinformatics for structural modelling

Another approach was the one presented by Luis M Escudero, Ramón y Cajal researcher at the University of Sevilla. He presented his research about the folding of curved epithelia, which lead to the discovery of a new geometrical shape named scutoid. The pressure on cells when curving the epithelium under other geometrical models is unsustainable. Nevertheless, when cells adopt a scutoid shape the distribution of tension is much more efficient. This was made by using mathematical and biophysical models using Voronoi diagrams. This new shapes have become viral in internet, and we can find even earrings shaped like our epithelium. Maybe we will see soon the integration of this shapes into new materials, more flexible and more resistant.

 

New reference software

During XIV Symposium on Bioinformatics a large variety of new software, in development or just finished, was presented. Among them, one in particular stood out for having become a global reference in single-molecule long reads sequencing, and has been included in the official pipeline of PacBio. SQANTI has been developed by Genomics of Gene Expression group, directed by Ana Conesa, with researchers from Centro de Investigación Príncipe Felipe and University of Florida. This software aims to become the «fastqc of long reads» as A Conesa said.

SQANTI pipeline starts from a reference genome, its annotation and the full transcripts obtained with PacBio. It outputs the corrected transcriptome, its annotation and a quality control report with 38 diagnostic plots. This allows us to study the sequenced transcriptome and identify new transcripts, or their isoforms. SQANTI allows us to implement machine learning to combine variables and obtain a curated transcriptome, removing isoforms potential to be artifacts.

 

 

SQANTI’s success is also due to its low false discovery rate (FDR), around 11.1%. Other filtering options for long reads have an FDR over 50%, which makes SQANTI a lot more reliable than its rivals.

Ana Conesa’s group also presented two other software: Tappas and PaintOmics.

 

National Bioinformatics Institute (INB)

Spanish National Bioinformatics Institute is a technological platform from Instituto de Salud Carlos III (ISCIII) with two main goals:

• to line up with ELIXIR goals (European platform to coordinate bioinformatics resources and big data related to life sciences) and increase synergies with this institution
• to promote the connection between translational medicine with the national health system

Although INB was found in 2003, it’s currently undergoing a restructuring and expansion process. The number of groups that participate as nodes and contribute with bioinformatics resources has increased from 10 to 19. It should be highlighted that four of nine new nodes are led by women. This reflects the inclusion of women in managerial positions in science, which supports the representation of women in this field that has a manifested gender bias. INB has also started TransBioNet, a network to encourage the translational capacity of Spanish bioinformatics community. Its aim is to support the consolidation of a network with bioinformatics groups and promote the training of professionals

INB’s web portal also offers information about future events relevant to national bioinformatics, like courses or conferences, and job lists to facilitate hiring bioinformatics professionals.