Large language models (LLMs) have recently gained significant attention as a promising approach to accelerate scientific discovery. However, their application in open-ended scientific domains such as biology remains limited, largely because they lack factually grounded and actionable explanations. To address this, we introduce a structured explanation formalism for virtual cells that represents biological reasoning as mechanistic action graphs, enabling systematic verification and falsification. Building on this, we propose VCR-Agent, a multi-agent framework that integrates biologically grounded knowledge retrieval with a verifier-based filtering approach to autonomously generate and validate mechanistic reasoning. Using this framework, we release the VC-TRACES dataset, which consists of verified mechanistic explanations derived from the Tahoe-100M atlas. We observed that training with these explanations improves factual precision and provides a more effective supervision signal for downstream gene expression prediction on TahoeQA. These results underscore the importance of reliable mechanistic reasoning for virtual cells, which can be enhanced through the synergy of a multi-agent system and rigorous verification.