The natural solution would be to use a aligner to map the probe sequences to the apple genome. I would use a splice-aware aligner although that might not be essential. You would have to create a FastQ file containing all the probe sequences. You would also have to download the apple genome from NCBI as well as gene annotation for the same species. Then align the probe sequences to the genome and assign each one if possible to a gene. The alignment and assignment can be done efficiently using the Rsubread package for example (using Unix rather than Windows).