CASE STUDIES

Pharmaceutical laboratory: identify nucleic acid-based oncology biomarkers (DNA and RNA) in liquid biopsies (plasma), with the aim of developing a non-invasive diagnostic tool and personalized therapy

Implementation of a protocol for the sequencing and identification of tumor biomarkers

Protocol writing: sample preparation, choice of suitable NGS (high-speed) technologies (Illumina, Roche 454, Ion torrent: Proton / PGM, SOLiD
Quality control / data cleansing
In silico analysis of RNASeq and DNASeq data:
– Expression calculation
– Comparison of expression across samples
– Exploratory data analysis (principal component analysis, multiple correspondence analysis)
Identification of differentially expressed gene signatures based on fold change, adjusted p-value and AUC (Area Under ROC Curve) (Volcano Plot, list of differentially expressed genes and Venn diagram of signatures)
Gene visualization (Heatmap, scatter plot). Statistical sorting of multi-criteria data
Functional gene analysis
– Inclusion of data from public and private databases
– Exploration of metabolic data
Statistical sorting of multi-criteria data (RNA, mutations, CNV, etc.) (Elastic Net, Random Forest)
Report writing, assistance with publication and presentation of results

Agrifood company: Analysis of seed genomics data to identify the genes involved in the germination process in the face of water stress in the context of a plant selection project

Carrying out the bioinformatics and biostatistics analysis stages of a genotyping project (identification of genetic and molecular signatures) for seeds.

Choice of sequencing strategy (illumina)
Analysis of NGS and microarray data (search for SSR microsatellites, SNP, insertion/deletion)
Analysis of NGS data
– Data mapping
– Identification of SSR microsatellites, SNP, insertion/ deletion
Quality control / data cleansing:
– Exploratory data analysis (PCA) enabling visual identification of a structure within the population (to then be factored into GWAS models)
– Data adjustment using a mixed model (with environmental effects as fixed and genotype effect as random)
Functional analysis of chromosomal events
Identification of differentially expressed gene signatures based on fold change, adjusted p-value.
Genome-wide association study (GWAS) enabling the analysis of genetic variations in several individuals in order to investigate the correlation between genotype and germination process
Analysis of linkage disequilibrium (LD) between markers
Visualization of p-values of genomic regions involved in character variation with a Manhattan plot
Writing of a report, assistance with publication and presentation of results

Cosmetics company: development of innovative cosmetic products

Analyze the impact of a cosmetic product on the skin using cutaneous microbiome data

Selection of the sequencing technology and analysis strategy: targeted 16S V3-V4 sequencing, Illumina
Implementation of analysis workflow
Data management: cross-referencing of metagenomic information with phenotype and environmental information
Identification of taxonomic profiles and microbial abundance of samples
– Alpha diversity (diversity, richness, variability indices)
– Beta diversity (distance calculations, PCoA)
Analysis at different levels of taxonomy: phylum, class, order, family, genus, (species)
Visualization interfaces (DataViz) and exploration of results

Agrifood company: selection of bacterial strains

Annotation of strain bank and identification of strains of interest

Hybrid assembly of NGS data
Structural and functional annotation of bacterial strains
Quality control
Comparative genomics
Characterization of operons of interest
Resistance and virulence research
Identification of metabolic pathways and prediction of enzymes of interest
Implementation of exploration interfaces

Pharmaceutical laboratory: fighting infectious diseases

Strain characterization and reconstruction of epidemiological history

Complete sequencing of bacterial strains
Mapping of data on the epidemiological reference genome
Identification of SNPs and chromosomal events
Characterization of variants (genes, functions)
Integration of specialized databases
Prediction of resistance and virulence
Phylogenetic classification

Veterinary laboratory: joint analysis of multi-source data in association with productive livestock farms

Creation of machine learning algorithms to detect genetic variations associated with phenotype data

Management of database infrastructure and calculation basis infrastructure
Creation of an algorithm/machine learning/deep learning
Building of an algorithm for imputation of missing data using various methods (machine learning, etc.)
Implementation of “benchmark” functionality for field teams with comparisons across individuals, products etc.
Creation of an application based on the algorithms developed and user training