But scientists won't see the signals if they're drowned out by too much noise. Those features might just signal the next breakthrough in health and medicine. That means they can continue chasing down genuine DNA features. Now, with just a couple new lines of code, scientists can get more reliable explanations out of powerful AIs known as deep neural networks. While we can't trust AI to be perfect, it turns out that sometimes we can't trust ourselves with AI either.Ĭold Spring Harbor Laboratory (CSHL) Assistant Professor Peter Koo has found that scientists using popular computational tools to interpret AI predictions are picking up too much "noise," or extra information, when analyzing DNA. Now, it's AI-generated pizza and beer commercials. DOI: 10.1186/s13056-3Īrtificial intelligence has entered our daily lives. c–e A colored box and a corresponding sequence logo of a known motif from JASPAR (with a corresponding ID) are shown for comparison. c–e A similar plot is made for a c DeepSTARR model trained to predict enhancer activity via STARR-seq data, d Basset model trained to make binary predictions of chromatin accessibility sites via DNase-seq data, and e CNN model trained to predict base-resolution read-coverage values from ATAC-seq data in PC-3 cell line. b An ensemble average saliency map is shown in lieu of ground truth (bottom row). The sequence logo of ground truth is shown for CNN-deep-exp for a synthetic data. a, b CNN-deep-relu trained to make binary predictions on a synthetic data and b ChIP-seq data for ATF2 protein in GM12878.
#Lzip atf2 Patch
Sequence logo of the uncorrected saliency map (top row), gradient angles at each position (second row), and corrected saliency map (third row) for a patch from representative test sequences.
0 kommentar(er)
