Abstract:A key challenge today lies in the ability to efficiently handle multi-omics data since such multimodal data may provide a more comprehensive overview of the underlying processes in a system. Yet it comes with challenges: multi-omics data are most often unpaired and only partially labeled, moreover only small amounts of data are available in some situation such as rare diseases. We propose MODIS which stands for Multi-Omics Data Integration for Small and unpaired datasets, a semi supervised approach to account for these particular settings. MODIS learns a probabilistic coupling of heterogeneous data modalities and learns a shared latent space where modalities are aligned. We rely on artificial data to build controlled experiments to explore how much supervision is needed for an accurate alignment of modalities, and how our approach enables dealing with new conditions for which few data are available. The code is available athttps://github.com/VILLOUTREIXLab/MODIS.
Abstract:One of the major challenges arising from single-cell transcriptomics experiments is the question of how to annotate the associated single-cell transcriptomic profiles. Because of the large size and the high dimensionality of the data, automated methods for annotation are needed. We focus here on datasets obtained in the context of developmental biology, where the differentiation process leads to a hierarchical structure. We consider a frequent setting where both labeled and unlabeled data are available at training time, but the sets of the labels of labeled data on one side and of the unlabeled data on the other side, are disjoint. It is an instance of the Novel Class Discovery problem. The goal is to achieve two objectives, clustering the data and mapping the clusters with labels. We propose extensions of k-Means and GMM clustering methods for solving the problem and report comparative results on artificial and experimental transcriptomic datasets. Our approaches take advantage of the hierarchical nature of the data.
Abstract:We consider the kernel completion problem with the presence of multiple views in the data. In this context the data samples can be fully missing in some views, creating missing columns and rows to the kernel matrices that are calculated individually for each view. We propose to solve the problem of completing the kernel matrices by transferring the features of the other views to represent the view under consideration. We align the known part of the kernel matrix with a new kernel built from the features of the other views. We are thus able to find generalizable structures in the kernel under completion, and represent it accurately. Its missing values can be predicted with the data available in other views. We illustrate the benefits of our approach with simulated data and multivariate digits dataset, as well as with real biological datasets from studies of pattern formation in early \textit{Drosophila melanogaster} embryogenesis.