Model developed by Sample et al: Human 5 prime UTR design and variant effect prediction from a massively parallel translation assay. Type in a sequence of 50 bases ("A","C","T", and "G") to make a prediction of ribosome loading.

Dataset Description: This model was trained on 280,000 of the 300,000-member eGFP library. The remaining 20,000 sequences were withheld for testing. No augmentations are recorded.


View API [+][-]

POST /api/predict/68

  data : string - Text string to upload as input.

  label : string - Most likely classification predicted by model.
  confidences : array (optional) - [{
    label : string - Classification for which confidence interval is being provided.
    confidence : number - Likelihood of classification as predicted by model.