SeqUnwinder results (simulateOverlap_sequnwinder_05092017)

SeqUnwinder version 0.1.2 run completed on: 2017/05/09 18:44:29 with arguments: --out simulateOverlap_sequnwinder_05092017 --threads 10 --debug --memepath /Users/akshaykakumanu/Softwares/src/bin/bin/ --geninfo mm10.info --seq /Users/akshaykakumanu/genome/mm10.fa --genseqs simulateOverlap.seqs --win 150 --mink 4 --maxk 5 --r 10 --x 3 --maxscanlen 15

Error on training data

Class TP Rate FP Rate Precision Recall F-Measure MCC ROC Area PRC Area
X#B 0.725 0.067 0.683 0.725 0.704 0.643 0.935 0.736
Y#A 0.628 0.063 0.665 0.628 0.646 0.578 0.911 0.712
X#A 0.645 0.074 0.637 0.645 0.641 0.568 0.906 0.718
Y#C 0.673 0.055 0.708 0.673 0.69 0.63 0.911 0.757
X#C 0.746 0.05 0.749 0.746 0.748 0.697 0.951 0.803
Y#B 0.626 0.082 0.604 0.626 0.615 0.536 0.896 0.678

Stratified cross-validation

Class TP Rate FP Rate Precision Recall F-Measure MCC ROC Area PRC Area
X#B 0.378 0.083 0.475 0.378 0.421 0.324 0.801 0.442
Y#A 0.499 0.075 0.572 0.499 0.533 0.448 0.846 0.579
X#A 0.609 0.118 0.509 0.609 0.555 0.459 0.862 0.575
Y#C 0.635 0.085 0.598 0.635 0.616 0.536 0.88 0.647
X#C 0.497 0.076 0.568 0.497 0.53 0.445 0.857 0.578
Y#B 0.583 0.123 0.487 0.583 0.531 0.429 0.857 0.551

Weighted K-mer models.

De novo motifs

SubClass/Lables De novo motifs
X#A rc
rc
X#B rc
X#C rc
Y#A rc
rc
rc
Y#B rc
rc
Y#C rc
rc
rc
A rc
rc
rc
rc
rc
B rc
rc
C rc
rc
rc
X rc
rc
rc
rc
rc
Y rc
rc
rc
rc
rc
rc
rc
rc

Label specific scores of de novo motifs

Extended version Compact version

Discriminative motifs. Try inputting these motifs into STAMP for validation.