Fig. 1: Machine learning strategy. | Oncogene

Fig. 1: Machine learning strategy.

From: A multi-faceted discovery strategy identifies functional antibodies binding to cysteine-rich domain 1 of hDKK1 for cancer immunotherapy via Wnt non-canonical pathway

Fig. 1

A The objectives of Twist machine learning platform are to obtain a higher yield of hit clones with highest affinity, greatest sequence diversity, and the best biophysical properties. B KNN based classifier trained based on previous SPR binding data. C CNN based neural network used to train on previously identified leads and NGS data to predict binders.

Back to article page