Understanding biological response to stimuli requires identifying mechanisms that coordinate changes across pathways. One of the promises of multi-omics studies is achieving this level of insight by simultaneously identifying different levels of regulation. However, computational approaches to integrate multiple types of data are lacking. An effective systems biology approach would be one that uses statistical methods to detect signatures of relevant network motifs and then builds metabolic circuits from these components to model shifting regulatory dynamics. For example, transcriptome and metabolome data complement one another in terms of their ability to describe shifts in physiology. Here, we extend a previously described linear-modeling based method used to identify single nucleotide polymorphisms (SNPs) associated with metabolic changes. We apply this strategy to link changes in sulfur, amino acid and lipid production under heat stress by relating ratios of compounds to potential precursors and regulators. This approach provides integration of multi-omics data to link previously described, discrete units of regulation into functional pathways and identifies novel biology relevant to the heat stress response, in addition to generating hypotheses.