Mitra Barzine

Postdoctoral Researcher in Bioinformatics May 2021 - Apr 2023

From Gamets to Birth, Institut Cochin

My work explores Institut Cochin gene expression and regulation at the maternal-fetal in human and mouse as a part of the Myestrom ANR project. I have applied the cellranger pipeline (installed on the IFB core cluster) on the raw fastq files generated from 10x Genomics’ single-nuclei RNA-seq sequencing of choriodecidua and placenta samples. I have then analysed and compared the gene expression between C-section and vaginal delivery with Seurat and muscat. I have identified differentially expressed genes, new markers and characterised new cell types. I have also determined each nucleus’s initial donors (maternal or foetal) with Souporcell and then explored cell-cell communication with CellPhoneDB and NicheNet, which resulted in hypothesising new mechanisms.
For this highly exploratory work, I have designed and generated various visualisations.

To save time, I have forked previous R packages to modify the code or completely implement new functionalities, e.g. UpSetR and ShinyCell. Further, with the latter package, I have created SnRNAseq self-service, a portable shiny application (based on the PortableApps framework) to overcome the lack of an appropriate shiny server.

I have contributed to interpreting the results, drafting publications, and providing bioinformatics support
and insights to the team, which mainly comprises wet-lab biologists. I have worked effectively bridging
the gap between bioinformatics and wet lab biology. During my mission, I presented my work at
conferences and have participated in public outreach programmes such as supervising two ”Apprentis
Chercheurs”.

Research Support Officer III in Bioinformatics Feb 2021 - Dec 2022

Centre for Molecular Medicine and Biobanking, University of Malta

My work was part of the TargetID project, which is an extensive effort to identify potential targets for
Malta treatments and means of identifying people more at risk. Within this project, I have re-analysed public data from COVID-19 patients. I have also performed the primary and secondary analyses of about 1000 blood samples, comprising cases, relatives and controls. These blood samples were collected, along with many other phenotypic and other biological measurements, during a previous project about Maltese Acute Myocardial infarction (MAMI).

I have designed and implemented a high-throughput RNA-Seq data analysis pipeline leveraging
high-performance computing infrastructure (SLURM) to allow large-scale data processing. My pipeline
relays on snakemake with shell, R, and python and uses conda environments. I have collaborated with biologists and geneticists to identify potential problems, improve workflows, and optimize data analysis.

I have also developed and maintained a shiny server to ease biologists’ visualisation of the transcriptomic data and its interpretation. Our multidisciplinary team collaboration has highlighted possible disease-causing mutations (ongoing validation).

I have trained and mentored junior bioinformaticians and predoctoral students during my mission. I
I have also prepared materials to present findings at international conferences and developed and taught training materials for various concepts and bioinformatics tools.

PhD candidate Life Sciences Oct 2013 - Dec 2020

Brazma research group, European Bioinformatics Institute

My doctoral research in the EMBL-EBI Functional Genomics research group explored the robustness of human gene expression extracted from independent transcriptomic (bulk RNA-Seq) and proteomics (label-free MS) data sets. To this end, I have completely reprocessed four RNA-Seq datasets in addition to the GTEx RNA-Seq data (QC, alignment, quantification, and normalisation) before applying various visualisation approaches and statistical analyses to explore the landscape of human gene expression baseline across tissues and studies. To gain valuable biological insights, I have performed many enrichment analyses (GO, pathways) and gained solid experience with querying and working with bioinformatics databases. My studies have highlighted sets of genes robustly expressed across tissues and studies at transcriptomic and proteomic levels. The tissue-specific genes and metabolic ones have exhibited the most robust expression. Besides the many analyses I have conceived to compare and integrate normal gene expression at RNA and protein levels, I have also devised a proteomic quantification method that increased the number of quantified proteins by including degenerated peptides. The complete set of tools I have developed for this thesis is distributed as an R package (barzinePhdR). In parallel, the gene expression of the different datasets (processed with the same pipelines) is distributed as another R data package (barzinePhdData). On another project, I helped with the validation of protein candidates. Finally, I have also taught many bioinformatic courses and modules. 

University of Cambridge 2013 - 2020

Field of study: Life Sciences
Degree: PhD

Thesis: Investigating Normal Human Gene Expression in Tissues with High-throughput Transcriptomic and PhD Life Sciences Proteomic Data [doi:10.17863/CAM.69592]

Defended on 1 December 2020.

Thesis advisor: Dr Alvis BRAZMA — Jury: Pr Kathryn LILLEY and Pr Jürg BÄHLER

Université de Nantes 2009 - 2011

Field of study: Bioinformatiques
Degree: Master II

Université de Nantes 2007 - 2009

Field of study: Chimie fine et Thérapeutique
Degree: Master II