Project Summary Collaborations Pharmaceuticals Inc. is a small business focused on developing treatments for difficult to treat diseases. The ongoing overreliance on opioids for chronic pain despite their poor ability to improve function has contributed to a significant and alarming epidemic of opioid overdose deaths and addictions in recent years. Despite the overwhelming addiction crisis, few therapies exist, and with low efficacy. Thus, a critical unmet need is an effective and safe treatment for opioid abuse disorders. Recent research has suggested that psychedelics are also capable of reducing drug-dependence, including opioid abuse, and that serious adverse effects are extremely rare. This drug abuse cessation is linked to the induction of neuritogenesis and increased neuroplasticity, a hallmark of psychedelics. While the psychedelic experience and neuroplasticity induction appear interlinked, several analogs of psychedelics have been proposed which induce neuroplasticity and drug- avoidance while seemingly lacking the psychedelic experience. Although few in number, these analogs have been named “psychoplastogens” and are promising candidates for treatment of opioid drug abuse. The de- coupling of the psychedelic experience and neuroplasticity induction is linked to receptor specificity as psychoplastogens are specific to 5HT2A. Here, we propose to use a combination of generative machine learning models with our extensive experience of supervised machine models to generate new analogs of psychedelics that will have improved 5-HT2A specificity, ADME properties, and are easily synthesizable to rapidly expand the pool of known psychoplastogens, significantly increasing potential therapeutic options for opioid abuse treatment.