Its true that selecting appropriate method(s) for statistical data analysis, based on a thorough understanding of how various factors produce measured outcomes, is vital to reaching correct research conclusions. I have had frustrating discussions with statisticians who did not understand biological mechanisms and with scientists who had very little experience with data analyses, but before attempting to develop your own statistical analysis tool (which would probably require multiple data sets to validate), have you met together with your supervisors and statisticians to jointly review your project data?