The functional genes underlying phenotypic variation and their interactions represent “genetic mysteries”. Understanding and utilizing these genetic mysteries are key solutions for mitigating the current threats to agriculture posed by population growth and individual food preferences. Due to advances in high-throughput multi-omics technologies, we are stepping into a biological Big Data era that is certain to revolutionize genetic research. In our lab, we devise methods for constructing and analyzing biological Big Data and demonstrate how it can be used as a versatile tool to dissect genetic mysteries. We implement an integrated strategy that could revolutionize genetic research by combining biological Big Data with machine learning, which involves mining information hidden in Big Data to identify the genetic models or networks that control various traits. Finally, we propose promising designed breeding strategies (CropGPTer=BT+IT+AT) utilizing the biological Big Data to improve crop yields and quality.