I was wondering if there is a way to run MRT analysis for a select two genes against a longer list of genes. This way, rather than every combination of genes to be considered from the list, every combination of genes for only the two select genes will be run.
Acronymitis alert! I guess you mean Multivariate Regression Trees? Please, as a general rule avoid or explain acronyms for the sake of getting question intelligible to everyone.
Since you tagged your question as music, I think you are referring to the "mutation relation test". Please avoid using acronyms like MRT in questions, and be as descriptive as possible. This reduces the odds that your question is glossed over and ignored.
music mutation-relation works only as you described. It tests every pairwise combination of genes in the provided --maf-file.
To test the mutation status of 2 genes against the statuses of several other genes, you can use music clinical-correlation instead. The mutation status of the 2 genes can either be provided in the --maf-file or in --input-clinical-correlation-matrix-file. The mutation status of the other genes can be a categorical clinical datatype, that you can fillup in matrix and pass into --categorical-clinical-data-file. This matrix can be generated by music mutation-relation using the --mutation-matrix-file argument. See the documentation.
Acronymitis alert! I guess you mean Multivariate Regression Trees? Please, as a general rule avoid or explain acronyms for the sake of getting question intelligible to everyone.