About myself
Welcome to my small site! My name is Sergi Picart and my background is Industrial Engineering and a Degree in Mathematics. I deeply thank CFIS (UPC, Barcelona, Spain) for offering me this excellent opportunity.
I used to be a PhD student in Biomedical Engineering, advised by Alexandre Perera in the B2SLab. The defense took place virtually in Covid19 times and can be found here. You will find further information about the publications and the software in the sections below.
I enjoy playing with math, statistical learning and programming. Likewise, computational biology is an amusing and rewarding field to explore. My main research topic included interpretable diffusion algorithms on biological networks, combined with functional analysis on metabolomics, transcriptomics and proteomics studies. I am currently employed by Boehringer Ingelheim in a small town in the south of Germany.
Contact
Publications
- Null diffusion-based enrichment for metabolomics data [Bioconductor page]
- FELLA: an R package to enrich metabolomics data [Bioconductor page] [OmicX] [slides]
- diffuStats: an R package to compute diffusion-based scores on biological networks [Bioconductor page] [OmicX] [slides]
- Benchmarking network propagation methods for disease gene identification [bioRxiv] [Repo]
- The effect of statistical normalisation on network propagation scores [bioRxiv] [Repo]
Publications from our collaborations
- MultiPaths: a python framework for analyzing multi-layer biological networks using diffusion algorithms [bioRxiv] [Repo]
- Non-targeted metabolomics reveals alterations in liver and plasma of gilt-head bream exposed to oxybenzone
- Amitriptyline at an environmentally relevant concentration alters the profile of metabolites beyond monoamines in gilt‐head bream
Repositories
Please visit our GitHub home page to find about our public software tools! My contributions include:
- FELLA [github] [bioconductor] [method] [software]: R package to enrich metabolomics data using knowledge graphs from the KEGG database.
- diffuStats [github] [bioconductor] [method] [software]: general purpose software to compute propagation scores, with or without statistical normalisation.
- Benchmark [github] [study] of network propagation algorithms to find disease genes
What should you do if you encountered an issue using those? In order:
- Look for your issue in a web search engine and, specifically, in the bioconductor forum (there is a number of answered questions there) and in the github issue list of the software package.
- If your question was not already asked but is general enough so that other people might benefit from it, please ask in one of the two locations stated above.
- Otherwise, drop me an e-mail with your query, ideally containing the name of the package in the subject line.
We appreciate you using our software and reaching us when in need!
FELLA databases NOT FOUND HTTP 404 error (04/10/2019) – Posted here since it is still being asked recurrently
Update: this issue was fixed in versions 1.4.2/1.5.2 (although I recommend from version 1.4.3 onwards). Make sure your FELLA version is updated, you can update and check via the commands below. If this does not work, please ensure you don’t use old versions of R/Bioconductor, or alternatively download and install from our github repository.
# install new one
BiocManager::install("FELLA")
# check version
packageVersion("FELLA")
Old solution: Our in-house R package FELLA is failing due to a recent problem with one of its dependencies, the KEGGREST package. KEGGREST is used to retrieve the annotations from KEGG, needed to build the knowledge graph of the desired species. We are waiting for a fix in KEGGREST, and to offer a temporary solution we are sharing some FELLA databases for common organisms, generated using the default parameters. The KEGG releases that were used to generate these range from 84.0+/10-08 to 90.0+/06-01, below you can find how to figure out which one corresponds to your object. Format: fella-keggorganismcode.zip
To use them, please download the zip file and unzip it to a directory of your choice. For example, unzipping fella-hsa.zip to ~/Downloads/fella-hsa:
library(FELLA)
FELLAdata <- loadKEGGdata("~/Downloads/fella-hsa", internalDir = FALSE)
# KEGG release
cat(getInfo(FELLAdata))
# ... and resume your analysis
Please be aware that these objects are distributed without the pre-computed matrices, otherwise they would become too heavy.