PLIS is a multiple testing procedure for testing several groups of hypotheses. Linear dependency is expected from the hypotheses within the same group and is modeled by hidden Markov Models. It is noted that, for PLIS, a smaller p value does not necessarily imply more significance because of dependency among the hypotheses. A typical application of PLIS is to analyze genome wide association studies datasets, where SNPs from the same chromosome are treated as a group and exhibit strong linear genomic dependency.
