mlp classifier (SoftMax Inc)
Structured Review

Mlp Classifier, supplied by SoftMax Inc, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
https://www.bioz.com/result/mlp classifier/product/SoftMax Inc
Average 90 stars, based on 1 article reviews
Images
1) Product Images from "GeoBind: segmentation of nucleic acid binding interface on protein surface with geometric deep learning"
Article Title: GeoBind: segmentation of nucleic acid binding interface on protein surface with geometric deep learning
Journal: Nucleic Acids Research
doi: 10.1093/nar/gkad288
Figure Legend Snippet: Overflow of GeoBind. ( A ) An illustration of the NBP surface segmentation. GeoBind takes the whole protein surface as input and outputs the segmented surfaces with each point assigned a likelihood of being involved in nucleic acid binding events. ( B ) Definition of binding interface and site. Points located on the protein surface with a distance to nucleic acids less than a cutoff are considered binding interfaces. Binding sites refer to residues that close to nucleic acids according to a similar definition. The mapping from an interface score to a site score is achieved through a max pooling operator. ( C ) Left, an introduction to quasi-geodesic convolution. We assign each point on cloud with a LRF. GeoBind makes use of quasi-geodesic distance and relative position (computed by LRF) for geometric embedding. Middle, the geodesic distance between two points is estimated by their position and orientation. Right, the relative position of a point in the LRF refers to the projection of the point on three axes. ( D ) Each point is initialized with three types of descriptors, namely multiple sequence alignment (MSA), chemical environment and curvature information. ( E ) Basic architecture of GeoBind. GeoBind consists of four blocks which contain the quasi-geodesic convolutional layers. In each block, a Multi-Layer Perceptron (MLP) is set up before and after the quasi-geodesic convolutional layer for the pre- and post-encoder.
Techniques Used: Binding Assay, Sequencing, Blocking Assay