[Geuvadis_rna_analysis] Geuvadis analysis group TC April 5th - minutes

Dear all, Here are the minutes of the Geuvadis analysis group TC of Thursday 5th of April. Apologies for the long email, it's been 2 hectic weeks of data processing and we went through a lot of things: Fastq files: - All the labs have uploaded mRNA and miRNA fastqs, except for 48 miRNA samples from Kiel where sequencing was delayed, and a few failed miRNA samples that will be skipped. Checksums are OK. - Kiel will sequence the missing miRNA samples next week - Thasso discovered that labs 2,3,6,7 produced fastq files without reads that fail Illumina quality checks, whereas 1,4,5 had them included - in order to be uniform, they should be filtered from all which has been done by Natalja & Tuuli. Two labs submitted 50bp miRNA reads and Tuuli clipped them. All the new files have been uploaded to the ftp site. - Next week the final set of filtered files, checksums and sample metadata will be put under clear ftp directories as the first raw data freeze - this will still be without more refined QC filters. mRNA data processing - GEM BAM files had format issues and need to be recreated. This is being run for the sandbox dataset, and Tuuli will check that this version is fine before rest of the data is processed. - In order to proceed with the analysis in the meanwhile, Tuuli is finishing mapping with bwa and will start submitting these bams too. The first eQTL and ASE analysis will be based on bwa. - All the mapping result files that were based on nonfiltered data from labs 1,4,5 need to be filtered afterwards - Tuuli and Thasso have taken care of this. - Micha will upload transcript quantifications (RPKM and read counts) to the ftp site, and Tuuli will upload exon read counts - these will happen within a week. - According to Ivo, they are planning to submit the GEM manuscript next week. miRNA data processing - Marc F. has downloaded all the fastqs and started running the first steps of their miRNA pipeline. The basic analysis should be done in 1-2 weeks. Genotypes - Of our 465 samples, 42 are not in 1000g Phase1 dataset and have to be imputed from 2.5M Omni genotypes. Natalja and Thomas W. from Munich had done this in parallel using Impute2. However, Tuuli detected a serious problem in the results - for some reason the imputation quality scores and the allele frequencies are very bad. The imputation will need to be debugged and redone. Tuuli, Thomas and Natalja will look into this next week. The good news is that we can do analysis for now using just the 423 Phase1 samples. Analysis group meeting in Geneva - 19 people will attending. Tuuli presented a draft of the program - first day brainstorming and presentations from all the labs, second day of building a concrete analysis plan with responsibilities and deadlines. Biology of Genomes meeting at Cold Spring Harbor - We got a poster presentation - which is not a bad thing considering the timeline, the meeting is 4.5 weeks. - We went briefly through some main analysis items that we can get done by then at least as first versions: eQTLs and their annotations, ASE, loss-of-function variants, splicing variation, quantitative/qualitative gene expression variation, miRNA variation and miRNA-mRNA correlation, unannotated transcripts... It will be a very nice poster if we work hard. Collaboration with Daniel McArthur (from Boston, previously at Sanger) - Dan is a loss-of-function specialist who has expressed interest in being a collaborator in Geuvadis. We already have analysts who will be working with LoFs and we should make sure that Daniel would do analysis that overlaps with this. However, his expertise in calling putative LoF variants from the 1000g data and pruning out false genotype calls could give us a good additional resource. People were generally positive about including him, and Tuuli will discuss with him in more detail what his interests and contribution would be. That's all for now - happy Easter to everyone! best regards, Tuuli -- Tuuli Lappalainen, PhD Department of Genetic Medicine and Development University of Geneva Medical School CMU / Rue Michel-Servet 1 1211 Geneva 4 Switzerland Tel. +41-(0)22-3795550 tuuli.lappalainen(a)unige.ch