Codes et logiciels

Statistics and machine learning

  • mixOmics: mixOmics offers a wide range of multivariate methods for the exploration and integration of biological datasets with a particular focus on variable selection. (GPL >=2)

  • GEMS-AI: Machine Learning Algorithms are trained on database to provide accurate predictions. But the main issue is how far can we trust them ? Can we certify the behaviour of these algorithms ? How can we be sure that the outcome of the algorithms will be reliable and unbiased ? This need for explanation of AI-based algorithms is at the core of GEMS-AI. (GPL3)

  • Jointseg: Methods for fast segmentation of multivariate signals into piecewise constant profiles and for generating realistic copy-number profiles. A typical application is the joint segmentation of total DNA copy numbers and allelic ratios obtained from Single Nucleotide Polymorphism (SNP) microarrays in cancer studies. The methods are described in Pierre-Jean, Rigaill and Neuvial (2015) <doi:10.1093/bib/bbu026>. (LGPL-2.1 | LGPL-3)

  • adjclust: Implements a constrained version of hierarchical agglomerative clustering, in which each observation is associated to a position, and only adjacent clusters can be merged. Typical use cases include Genome-Wide Association Studies or Hi-C data analysis, where the similarity between items is a decreasing function of their genomic distance. The implemented algorithm is time and memory efficient, see Ambroise et al (2019) <doi:10.1186/s13015-019-0157-4>. (GPL-3)

  • sanssouci: The goal of sanssouci is to perform post hoc inference: in a multiple testing context, sanssouci provides statistical guarantees on possibly user-defined and/or data-driven sets of hypotheses. Typical use cases include differential gene expression (DGE) studies in genomics, and fMRI studies in neuroimaging. (GPL-3)

  • coordinate layout: Plugin for cytoscape 3.0 to explicitly represent the graph nodes at pre-defined coordinates.

  • LargeGraphCenterEstimator: Python package to estimate the barycenter of large weighted graphs.

  • CoSeq: Bioconductor package for high-throughput sequencing data analysis.

  • SelVarMix: R package for variable selection in supervised and unsupervised classification.

  • MaskMeans: Multi-view agglomeration/splitting K-means clustering algorithm.

  • W2reg: Train neural networks classifiers (in PyTorch) with Wasserstein-2 regularization to reduce algorithmic bias in future predictions

Signal and Image processing

  • FitEllipsoid: Java plug-in for ICY (free software). Image processing and ellipsoid delineation.

  • Utilzreg: Image registration algorithms for 2D and 3D images. Includes implementations of the "LDDMM", "Geodesic Shooting" and "LogDemons" algorithms. C++ implementation with OpenMP parallelisation.

  • Geoshoot: Diffeomorphic image registration using the Geodesic Shooting approach. C++ code with GPU parallelisation in OpenCL. (MIT License)

  • Fast Transform Learning: Matlab package to illustrate the effectiveness of a deep representation method for processing images of cancer in a dozen organs.

  • Regularized non-local total variation: Matlab Package for image restoration.

  • SlicerRegularizedFastMarching: Slicer3D package for semi-interactive 3D image segmentation.

Numerical simulation

  • GPELAB: Matlab Toolbox for the simulation of quantic systems.

  • Getfem++: Finite element library in C++. Contribution to Fluid (Navier-Stokes), fictitious domain methods, domain decomposition (Parallel implementation).

  • DassFlow-Py: Python code for estimating bedrock topography under polar ice caps (Antarctica).