Wednesday, 7th May, 2014, 11:00

Understanding alternative splicing by genome-wide, quantitative profiling

Both chromatin state and binding of splicing factors to regulatory sequence elements of pre-mRNA have been shown to affect alternative splicing outcomes.  Yet the precise interplay between these two determinants is not known.  The availability of a relatively large number of relevant, publicly available, high-throughput ChIP-seq, CLIP-seq, and RNA-seq datasets make it possible to study this in depth on a genome-wide scale.  Read profiling is a commonly used method to increase the signal strength of high-throughput data by combining reads from a set of similar loci rather than examining each locus individually, thus increasing the signal and statistical power.  However, profiles often convey only qualitative information,  In this talk I will present a method we have developed to calculate exact P-values for comparison of a profile with a proper control.  The method allows for single-nucleotide resolution in principle, and can be used on most types of high-throughput sequenceing data.  I will also show how we have applied the method thus far to study the relationship between chromatin and splicing factors.

Speaker: Isaac J. Kremsky Computational Genomics group, GRIB

Room Aula

Sunday, 27th April, 2014

GPCR Spring Conference 2014, Selent, Jana i Martí, Maria. Comité organitzador.

PRBB (Barcelona) 28-30/4/2014. 

Thursday, 13th March, 2014, 11.00-12.00

What can the cloud do for you? Solving health puzzles through cloud computing.

Recent years, the use of computer-aided drug discovery techniques has been increased. These techniques, that includes, for example, virtual screening, compound selection, activity prediction and modelling, allow us to optimize our research work, reducing time and money. But its usage requires software (usually commercial and expensive) and high HPC (high performance computing) systems, which implies to invest money in computational infrastructure.  Cloud solutions for drug discovery helps us to "optimize our optimization process". Cloud systems are elastic, requires less money investment and are always ready, paying only for the resources you are really using.

Speaker: Alfons Nonell-Canals, PhD, CEO Mind the Byte, S.L.

Room Xipre Room 173.06 - PRBB

Wednesday, 26th February, 2014, 11:00

Are long non-coding RNAs translated?: Ribosome profiling reveals the complexity of eukaryotic proteomes

We have compiled raw sequencing data from ribosome profiling experiments performed in different species (human, mouse, zebrafish, fruit fly, yeast) and used them to assemble transcripts, quantify transcript-ribosome associations, and investigate the coding potential and strength of purifying selection of the putatively translated open reading frames in some long non-coding RNAs. We detected extensive association of lincRNAs with ribosomes, not only observed in mammals but also in the other eukaryotic groups. Surprisingly, some of the lncRNAs show significant sequence similarity to proteins only annotated in Genbank, whereas others show not such similarity but still contain putative short open reading frames. The coding potential of ribosome-associated lncRNAs ORFs, measured using different codon frequency based sequence statistics, is intermediate between intronic ORFs and experimentally validated ORFs. These ORFs are subject to weaker selective constraints than most experimentally validated proteins as inferred from single nucleotide polymorphism densities. Our results suggest that many lncRNAs have coding properties and that this class of genes most likely includes protein-coding genes that are no longer functional as well as genes encoding new, poorly-conserved, peptides.

Speaker: Jorge Ruiz Orera - Evolutionary Genomics, GRIB (IMIM - UPF)

Room Aula-4th floor

Sunday, 16th February, 2014, 11:00

Rethinking Binding and Binding Kinetics

Speaker: Jose Duca, Head of Computer-Aided Drug Discovery at Novartis

Room Marie Curie Room

Wednesday, 12th February, 2014, 11:00

Alternative splicing patterns in the Cancer Genome Atlas datasets

Speaker: Endre Sebestyén - Computational Genomics, GRIB (IMIM-UPF)

Room Sala 473.10

Tuesday, 21th January, 2014, 12:00

Integrative strategies for biomedical knowledge.

The progress in biomedical research is increasingly hampered by the difficulty of managing and jointly exploiting the huge amount of information that is accumulated and that it is often fragmented into silos. Consequently, the biomedical knowledge discovery can deeply benefit from scientific, technological and organisational advances in the ways that biomedical information is integrated and jointly analysed. Examples of efforts that we are doing in this direction are: 1) the IMI eTOX project, which tries to advance in the in silico prediction of the potential in vivo toxicity of drug candidates by means of information sharing among the pharmaceutical companies and the application of multi-level modelling strategies; 2) DisGeNET, which is a gene-disease database created by integration of gene-disease associations from several resources; 3) the EU-ADR project, which has developed innovative pharmacovigilance strategies by means of joint exploitation of millions of European healthcare records followed by bioinformatics substantiation of the drug-event signals detected; and 4) the recently started IMI EMIF project, which aims to develop a common information framework of patient-level data that will link up and facilitate access to diverse medical and research data sources, opening up new avenues for research.

Speaker: Ferran Sanz, professor of Biostatistics and Biomedical Informatics at Pompeu Fabra University (UPF) and Director of the IMIM-UPF Research Programme on Biomedical Informatics (GRIB).

Room PRBB Auditorium

Wednesday, 8th January, 2014, 11:00

“Analyzing the exome of 7000 tumors to identify cancer drivers and their therapeutical opportunities”

Speaker: Carlota Rubio - Biomedical Genomics, GRIB

Room Aula (473.10)

Sunday, 8th December, 2013

DisGeNET RDF and its implementation in Open PHACTS

SWAT4LS 2013 – International Semantic Web Applications and Tools for Life Sciences Workshop in Edinburgh (UK)

Speaker: Nuria Queralt - Integrative Biomedical Informatics group (GRIB)

Wednesday, 27th November, 2013, 11:00

"Alteration of alternative splicing and RNA binding proteins across multiple tumors"

Current cancer genomics projects apply high-throughput technologies to discover recurrent genetic variations in patient samples. These efforts are crucial to describe the genetic diversity of cancer and to classify into novel subtypes for improved prognosis and therapeutics. These genome-scale studies focus mostly on the detection of alterations of the DNA and the expression of genes. However, alterations in Alternative Splicing (AS), which hold important signatures that can provide novel prognostic and therapeutical strategies, have not been yet thoroughly characterized. We have used RNA sequencing data from The Cancer Genome Atlas (TCGA) project for hundreds of tumor samples and paired normal tissues to study the splicing changes and differential expression of Splicing Factors (SFs) and RNA binding proteins (RBPs) in 13 different cancer types. We find new proteins and splicing events that are recurrently altered in tumors. Additionally, by using correlations, we detect possible association of RBPs and events, suggesting splicing regulatory modules. Our analysis indicates that different cancers present similar alterations, suggesting a general path for cells towards cancer by concurrent splicing alterations. This analysis provides useful information to elucidate the impact of alternative splicing in the functional dynamics of cell transformation in multiple cancer types and may help uncovering novel therapeutic strategies.

Speaker: Eduardo Eyras - Computacional Genomics group of GRIB (IMIM - UPF)

Room Aula (473.10)