GMEmbeddings: An R Package to Apply Embedding Techniques to Microbiome Data

Abstract

Large-scale microbiome studies investigating disease-inducing microbial roles base their findings on differences between microbial count data in contrasting environments (e.g., stool samples between cases and controls). These microbiome survey studies are often impeded by small sample sizes and database bias. Combining data from multiple survey studies often results in obvious batch effects, even when DNA preparation and sequencing methods are identical. Relatedly, predictive models trained on one microbial DNA dataset often do not generalize to outside datasets. In this study, we address these limitations by applying word embedding algorithms (GloVe) and PCA transformation to ASV data from the American Gut Project and generating translation matrices that can be applied to any 16S rRNA V4 region gut microbiome sequencing study. Because these approaches contextualize microbial occurrences in a larger dataset while reducing dimensionality of the feature space, they can improve generalization of predictive models that predict host phenotype from stool associated gut microbiota. The GMEmbeddings R package contains GloVe and PCA embedding transformation matrices at 50, 100 and 250 dimensions, each learned using ∼15,000 samples from the American Gut Project. It currently supports the alignment, matching, and matrix multiplication to allow users to transform their V4 16S rRNA data into these embedding spaces. We show how to correlate the properties in the new embedding space to KEGG functional pathways for biological interpretation of results. Lastly, we provide benchmarking on six gut microbiome datasets describing three phenotypes to demonstrate the ability of embedding-based microbiome classifiers to generalize to independent datasets. Future iterations of GMEmbeddings will include embedding transformation matrices for other biological systems. Available at: https://github.com/MaudeDavidLab/GMEmbeddings.

Deep learning methods in metagenomics: a review

“…Once the matrix of embeddings is created, the new data are simply multiplied by the embedding matrix to produce a new table of embedded data. GMEmbeddings [ 145 ] provides embeddings based on GloVe [ 146 ], an NLP algorithm, by aligning requested samples to known amplicon sequence variants (ASVs) using blast . This same GloVe algorithm can generate an embedding of a user-uploaded abundance matrix [ 147 ].…”

Deep learning methods in metagenomics: a review

“…The latter one produces a very smaller amount of sequences. This is why this data source is used by methods that focus on speed, such as [51] and [52], or methods that aim to create fixed pre-trained embeddings, like [53]. However, today, most methods for metagenomic analysis rely on Next-Generation Sequencing.…”

“…The goal of this approach is to improve the accuracy of disease prediction compared to using a single deep learning model. GMEmbeddings ( [53]) provides embeddings based on GloVe ( [127]), a Natural Language Processing algorithm, by aligning requested samples to known ASV using BLAST. ( [128]) uses the same GloVe algorithm to generate an embedding of a user-uploaded abundance matrix.…”

“…GMEmbeddings ([55]) provides embeddings based on GloVe ([129]), a Natural Language Processing algorithm, by aligning requested samples to known ASV using BLAST. ([130]) uses the same GloVe algorithm to generate an embedding of a user-uploaded abundance matrix.…”