Parameters
If you want to test the sample, you just need to upload the file and click the bottom start button.

HSM6AP

Tips:The web can only predict sequence. If you want to use our complete model, please contact us for the complete code. EP1 represents the Full transcript model; EP2 represents the mature mRNA model. Please comply: the name of the sequence uploaded cannot be repeated in the file. We recommend that users upload less than 1000 data at a time. If there is no feedback or the page is lost, please try to reduce the amount of samples appropriately.

We have presented HSM6AP, an predictor to distinguish M6A sites . Our webserver contains two models, which were trained from full Transcript and mature mRNA and were called Full and Mature, respectively. Based on the species of test dataset, the users can choose the model they want.

In this paper, sequence derived features and gene derived features are extracted and selected. Considering the possibility of cross-species prediction, our web only provides the support for predicting sequentially derived features. In fact, our final model can only predict methylation sites of humans, and require users to provide sequence information and genomic coordinate information.

If you need to predict human methylation sites and have both sequence information and genomic coordinate information, you can download our code to do that.

  • Sequence-based features:

       NCP, DNC, CKSNAP, Mismatch.

  • Physicochemical properties:

       PCPseDNC, PCPseTNC, PCPseTNC, PCPseTNC.

  • Genome-derived characteristics:

       eCLIP_HNRNPC_gr, YTHDC1_TREW_gr, YTHDF1_TREW_gr, YTHDF2_TREW_gr, miR_targeted_genes_grl, TargetScan_hg19_gr, verified_targets_gr, METTL3_TREW, METTL14_TREW, WTAP_TREW, METTL16_CLIP, ALKBH5_PARCLIP, FTO_CLIP, FTO_eCLIP.

  • If you have any problems,please contact me.

    lijingtju@foxmail.com