Patient diversity and unknown disease cause are major challenges for drug development and clinical trial design for ALS. Moreover, the heterogeneity of the ALS patient population is not reflected in currently available transgenic animal models making direct translation of potential therapeutics difficult. To address this, we utilize direct conversion technology to transform skin biopsies from ALS patients into neuronal progenitor cells (NPC). Using induced astrocytes (iAs) differentiated from these NPCs in co-culture with mouse embryonic motor neurons, we developed an in vitro ALS model to screen potential therapeutics. We have screened numerous compounds on multiple sporadic (sALS) and familial (fALS, mtSOD1 and C9ORF72) iAs and observed a diverse patient response to different therapeutic agents. Here, we utilize the compound, CuATSM, to subgroup therapeutic responsiveness of patient lines and identified shared dysfunctional pathways between responders. Following detailed analysis of ALS disease markers, elevated mitochondrial activity states was present in all CuATSM responders and was nonexistent in nonresponders. Treatment of iAs with CuATSM restored this activity to healthy control levels. Together, these findings suggest that iAs can be used to identify both disease modifiers and pathways dysregulated in an individual. Thus enhanced understanding of cellular profiles could facilitate personalized medicine based treatments.