Predictor

Predictor: a web server for predicting the 4acC modification possibility in Arabidopsis thaliana using a Weakly Supervised Model.

Condition:

Motif requirement:

The speed of motif outcome:

Input Data (Standard FASTA): *Note: Please input up to 10 DNA sequences of at least 200nt long each, using standard FASTA format as example shows.

Upoad FASTA File：

Available Options:

Condition: Four data conditions can be selected for the 4acCPred: standard - Arabidopsis gDNA (wild type), NH₂OH - Arabidopsis gDNA treated with NH₂OH samples, met - Arabidopsis gDNA treated with met1 mutant, and rdd- Arabidopsis gDNA treated with ros1dml2dml3(rdd) mutant.
Motif requirement: Users can choose whether 4acCPred needs to identify motifs (pattern sequences) of the data or not. If users choose 'require', 4acCPred will output a motif (pattern sequence) of the input sequence with the highest probability as well as the probability is larger than 0.5.
Speed: We provides two kinds of IG method to identify motifs (pattern sequence) of the input sequence with the highest probability (greater than 0.5).
- Fast mode: 4acCPred will implement zero matrixes and fixed letter frequencies (GC content) as a reference input import in the model.
- Slow mode: 4acCPred will construct a reference sequence by shuffling the original input while do not change the frequency of dinucleotides, which considers the real situation in the calculation because the dependability of integrated gradients (IG) depends on reference input in the model.