====== Florent GUINOT ======

{{:members:fguinot:photo.jpeg?150 |}} 
**Phd Student (CIFRE contract) with UEVE and Bioptimize** \\
Laboratoire de Mathématiques et Modélisation d'Evry (LAMME) \\ Equipe Stat & Génome \\
23 boulevard de France 91000 Evry\\
<guinotflorent@gmail.com> \\
{{:members:fguinot:cv_guinot.pdf|}} \\
[[https://drive.google.com/file/d/1hvQMRuGFDXRTO3t-EuqNwP0gtp-J9H4t/view?usp=sharing|Soutenance]] \\
[[https://fguinot.github.io/thesis_bookdown/|Manuscrit thèse]]

===== Thesis =====

** Statistical learning for omics association and interaction studies based on blockwise feature compression. **
  * Improvement of the statistical power in GWAS by taking in account the genome structure through combination of hierarchical clustering and penalized regression. 
  * Detection of large-scale interactions on structured datasets. Application for the research of interaction between genome and microbiome.

**Direction:** [[http://www.math-evry.cnrs.fr/members/Cambroise/welcome|Christophe Ambroise]], [[http://web4.ensiie.fr/~szafranski/|Marie Szafranski]]
 
===== Research topics =====
  * Statistical learning applied to high-dimensional data (omics data).
  * Hypothesis testing.
  * Genome-Wide association studies and Metagenomics.

===== Software =====
  * R package [[https://github.com/fguinot/sicomore-pkg|SIComORe]] : From a set of input matrices and phenotype related to the same set of individual, sicomore is a two-step method which (1) finds and select groups of correlated variables in each input matrix which are good predictors for the common phenotype; (2) find the most predictive interaction effects between the set of data by testing for interaction between the selected groups of each input matrix.

  * Webserver tool [[http://stat.genopole.cnrs.fr/leos|LEOS]]:  The LEOS web tool allows to perform Genome-Wide analysis with group of variables through hierarchical SNP aggregation. The program requires a matrix *X* of SNP coded as [0,1,2], a binary phenotype vector Y coded as [0,1] and optionally a genetic map indicating to which chromosome belongs the SNP. In order to work properly, the program requires that the columns *X* are ordered according to the SNP position on the genome. The program provides a manhattan plot indicating which clusters of SNPs are significantly associated with the phenotype, a table of results and a plot showing at which level in the hierarchy the SNPs have been aggregated. 


===== Publications =====
^           2018           ^^
| [1] | <color #705441>**Guinot, F.**, Szafranski, M., Ambroise, C. & Samson, F. </color> Learning the optimal scale for GWAS through hierarchical SNP aggregation. 2017. BMC bioinformatics. https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-018-2475-9 |
| [2] | <color #705441>**Guinot, F.**, Szafranski, M., Chiquet, J. & Ambroise, C.</color> Fast Computation of Genome-Metagenome Interaction Effects, 2018. (Preprint). https://arxiv.org/abs/1810.12169|
| [3] | <color #705441>**Guinot, F.**, Szafranski, M., Chiquet, J. & Ambroise, C.</color> Une approche hiérarchique de la recherche d’interactions entre données omiques, 2018. (national conference paper) https://toltex.u-ga.fr/users/RCqls/Workshop/jds2018/resumesLongs/subm356.pdf|


===== Teaching =====
  * TD Probabilités - Licence 1 Biologie (S2 2017/2018)
  * TP Probabilités / Introduction au logiciel R - Licence 1 Biologie (S2 2017/2018)
  * TD Probabilités - Licence 1 Biologie (S2 2016/2017)