Precioso, F. | Search bioRxiv

bioRxiv2024

scMusketeers: Addressing imbalanced cell type annotation and batch effect reduction with a modular autoencoder

The growing number of single-cell gene expression atlases available offers a conceptual framework for improving our understanding of physio-pathological processes. To take full advantage of this revolution, data integration and cell annotation strategies need to be improved, in particular to better detect rare cell types and by better controlling batch effects in experiments. scMusketeers is a deep learning model that optimises the representation of latent data and solves both challenges. scMusketeers features three modules: (1) an autoencoder for noise and dimensionality reductions; (2) a focal loss classifier to enhance rare cell type predictions; and (3) an adversarial domain adaptation (DANN) module for batch effect correction. Benchmarking against state-of-the-art tools, including the UCE foundation model, showed that scMusketeers performs on par or better, particularly in identifying rare cell types. It also allows to transfer cell labels from single-cell RNA sequencing to spatial transcriptomics. With its modular and adaptable design, scMusketeers offers a versatile framework that can be generalized to other large-scale biological projects requiring deep learning approaches, establishing itself as a valuable tool for single-cell data integration and analysis.

bioinformatics↗

bioRxiv2020

Towards CNN Representations for Small Mass Spectrometry Data Classification: From Transfer Learning to Cumulative Learning

Rapid and accurate clinical diagnosis of pathological conditions remains highly challenging. A very important component of diagnosis tool development is the design of effective classification models with Mass spectrometry (MS) data. Some popular Machine Learning (ML) approaches have been investigated for this purpose but these ML models require time-consuming preprocessing steps such as baseline correction, denoising, and spectrum alignment to remove non-sample-related data artifacts. They also depend on the tedious extraction of handcrafted features, making them unsuitable for rapid analysis. Convolutional Neural Networks (CNNs) have been found to perform well under such circumstances since they can learn efficient representations from raw data without the need for costly preprocessing. However, their effectiveness drastically decreases when the number of available training samples is small, which is a common situation in medical applications. Transfer learning strategies extend an accurate representation model learnt usually on a large dataset containing many categories, to a smaller dataset with far fewer categories. In this study, we first investigate transfer learning on a 1D-CNN we have designed to classify MS data, then we develop a new representation learning method when transfer learning is not powerful enough, as in cases of low-resolution or data heterogeneity. What we propose is to train the same model through several classification tasks over various small datasets in order to accumulate generic knowledge of what MS data are, in the resulting representation. By using rat brain data as the initial training dataset, a representation learning approach can have a classification accuracy exceeding 98% for canine sarcoma cancer cells, human ovarian cancer serums, and pathogenic microorganism biotypes in 1D clinical datasets. We show for the first time the use of cumulative representation learning using datasets generated in different biological contexts, on different organisms, in different mass ranges, with different MS ionization sources, and acquired by different instruments at different resolutions. Our approach thus proposes a promising strategy for improving MS data classification accuracy when only small numbers of samples are available as a prospective cohort. The principles demonstrated in this work could even be beneficial to other domains (astronomy, archaeology...) where training samples are scarce.

bioinformatics↗