The language with which microorganisms communicate with each other and with us is chemistry, with small molecule secondary metabolites being the alphabets of this language. The secondary metabolites modulate the dynamics of our microbial make-up and have been correlated to a number of diseases including infections, inflammation, cancer, and neurological disorders, yet it remains a challenge to identify small molecular effectors of human diseases in the clinically relevant conditions and their associated biological functions. We employ mass spectrometry-based approaches to identify and visualize metabolomic profiles of pathogens correlating with capacity to cause disease, for use in assessing disease status and for developing new treatment strategies. Herein, a combination approach based on molecular networking, MS2LDA, statistical models, and in silico annotation tools DEREPLICATOR, SIRIUS with CSI:FingerID and XCMS Online will be presented that enables visualization and dereplication of key phenotypically relevant molecules underlying infections by Burkholderia sp. bacteria in cystic fibrosis patients. Using this approach, we will highlight the personalized response of three Burkholderia sp. bacteria, B. cenocepacia, B. multivorans, and B. dolosa, isolated from cystic fibrosis patients to antibiotic trimethoprim. Specifically, we identified a new specie specific bacterial pathway for metabolism of antibiotic trimethoprim, strain-specific differential expression of known secondary metabolites such as fragin, pyrazines, ornibactin, diffusible signal molecules, N-acylhomoserine lactones and their analogs, and discovered new secondary metabolites that are produced in response to trimethoprim exposure. Together these results highlight the importance of investigating presence of personalized bacterial metabolomes in infectious diseases.