Bipolar disorder (BD) is known to be comorbid with many medical and psychiatric conditions and they may share inflammatory and stress-related aetiologies, which could give rise to this association. Psychosomatic illnesses are characterized by widespread immune signaling in the central nervous system, oxidative stress, and increased immune trafficking into the brain. The integrated stress response (ISR) responds to a variety of different stress conditions that lead to alterations in cellular homeostasis. However, as a causative mechanism underlying the cognitive deficits and neurodegeneration in a broad range of brain disorders, the impact of ISR on BD is understudied, particularly for the ISR’s central regulatory switch lies in eIF2 ternary complex formation. We aimed to prioritize eIF2-associated genes for follow-up studies using the summary data-based Mendelian Randomization (SMR) and Bayesian colocalization (COLOC) methods to integrate the summary-level data of the GWAS on BD and the expression quantitative trait locus (eQTL) study in blood. We utilized the GWAS data including 41917 BD cases and 371549 controls from the Psychiatric Genomics Consortium and the eQTL data from 31,684 participants of predominantly European ancestry from the eQTLGen consortium. Plus, we use Graph DB to alt drug targets. The SMR analysis identified EIF2B5 gene that were associated with BD due to no linkage but pleiotropy or causality. The COLOC analysis strongly suggested that EIF2B5 and the trait of BD were affected by shared causal variants, and thus were colocalized. Utilizing data in EpiGraphDB we find other putative causal BD genes (EIF2AK4 and GSK3B) to prioritise potential alternative drug targets.