Fig. 1: Workflow for sample collections, data processing, analysis, and machine learning model construction.
From: Harnessing vaginal inflammation and microbiome: a machine learning model for predicting IVF success
The scheme illustrates the model’s workflow. Swab samples were collected at 3 time points of the IVF cycle for data preprocessing of vaginal microbiome and inflammation(top). Vaginal microbiome was analyzed with 16S rRNA sequencing, and inflammation was analyzed with Luminex. Data was then used as a feature set, extracted and reduced before training an SVM model, which was cross-validated with leave-one-subject-out methods to construct a pregnancy prediction classifier with explanatory features. Created with BioRender.com.
