Microbiome & multi-omics / O2PLS
Your data stays in your browser: files are read and processed on your device, never uploaded to any server, and cleared when you close the page. Free, no sign-in.
O2PLS splits the variation of two data sets matched by sample (e.g. transcripts and metabolites) into three parts: joint variation shared by both, variation specific to X and variation specific to Y. A larger joint R² means more shared signal; “R²Y predicted from X” shows how much of Y the X joint components explain.
On the score plot each point is a sample (X joint scores); on the loading plot each point is a variable. X and Y variables pointing in the same direction are linked through the joint components; the further from the origin, the larger the contribution.
You choose the component numbers: this is a simplified O2PLS without cross-validation; typically 1–3 joint and 0–2 specific components. Too many components overfit. Keep “Scale variables” ticked when variables are on different scales.
Variables with missing values are removed and constant variables ignored. Samples are matched by name; only samples present in both tables are used.
Trygg J. & Wold S. (2003) Journal of Chemometrics 17:53–64; algorithm steps follow the o2m procedure of OmicsPLS (Bouhaddani S. el et al. 2018, BMC Bioinformatics 19:24), without cross-validation.
Up to 5,000 samples per table.
Send us your data and research question and you will receive a written plan within 1 working day: analysis steps, parameter rationale, deliverables and timeline. Quoted per project.