MINT algorithm explanations


I’m student in data sciences in Belgium.

I have read the paper you published on MINT (" MINT: a multivariate integrative method to identify reproducible molecular signatures across independent experiments and platforms"), but I don’t understand, in the first lines of the pseudo-code of the algorithm, why it is written that X^{(m)} = X (and same for Y). For me, X^{(m)} is just one bloc and X is the all dataset, is it wrong ?

Could you explain please ?

Thanks you


hi @lpanneel

There must be a typo here, thanks! it should be X = [X_1^{(1)}, …, X_1^{(m)},]^T so that we can then calculate the global components / loadings on the whole concatenated data matrix.
Maybe this picture can help to represent X and Y (left hand side)