A Dual-Filter Strategy Integrating CRISPR-based Target Screening and Text Mining for Hand-Foot Syndrome

The experimental high-throughput screening (HTS) methods, exemplified by CRISPR-based screening, have revolutionized target identification in drug discovery. However, such screens frequently yield extensive and unrelated target lists necessitating costly and time-intensive experimental validation. Here, we propose a dual-filter strategy that integrates literature-mined targets with CRISPR/Cas9 screening outputs, systematically prioritizing the most credible candidates and thereby reducing the experimental validation burden and increasing success rate. To validate this strategy, we applied it with hand-foot syndrome (HFS), a clinically challenging side effect induced by fluoropyrimidine treatment. We identified ATF4 as a key regulator of 5-fluorouracil (5-FU) toxicity in the skin and revealed forskolin as a potential therapeutic agent of HFS through the strategy. Mechanistically, forskolin triggers MEK/ERK-dependent ATF4 induction, subsequently driving 5-FU detoxification via the ATF4-mediated eIF2α/IκB signaling pathway. Our findings demonstrate that this dual-filter strategy could notably accelerate drug discovery by reducing experimental validation burden after target screening.