AlphaFind v2: Similarity Search in AlphaFold DB and TED Domains across Structural Contexts 1 AlphaFind v2 is a web application designed for fast structure-based similarity search in the AlphaFold Database of predicted protein structures, addressing the computational challenges of large-scale 3D structure comparison. 2 It combines fast pre-filtering using protein embeddings that preserve structural information with refinement via US-align, balancing search speed and biological relevance. 3 The tool offers six complementary search modes, including full-protein chain search, pLDDT-filtered searches at 70%, 80%, and 90% thresholds, TED domain search, and TED multidomain search. 4 It supports optional filtering by organism, taxonomy ID, or CATH label, and links search results to corresponding experimental protein structures. 5 The approximate search phase delivers results in seconds, with full structural refinement completed in under a minute on average, outperforming tools like FoldSeek Server and Merizo-search in both speed and average TM-Score. 6 Key applications include identifying homologous proteins in disordered regions via pLDDT filtering and detecting conserved multidomain architectures as demonstrated in case studies of PIN3 and NCAM1 proteins. 7 The web server is built with a Python backend, Flask REST API, Celery asynchronous tasks, OpenSearch vector database, and Kubernetes deployment for scalable performance. 8 AlphaFind v2 uses AlphaFold DB version 4 and precomputed embeddings, with all functionality freely accessible to users without login requirements. 📜Paper: biorxiv.org/content/10.64898… #AlphaFind #ProteinStructure #StructuralBiology #AlphaFold #Bioinformatics #StructuralSimilarity

J Chem. Inf. Model. KiSSim: Predicting Off-Targets from Structural Similarities in the Kinome @JCIM_JCTC #StructuralSimilarity #protein #proteinstructures @ChariteBerlin pubs.acs.org/doi/10.1021/acs…

#RSHighlyCitedPaper #RoadExtraction by Using Atrous Spatial Pyramid Pooling Integrated #Encoder-#Decoder Network and #StructuralSimilarity Loss By Hao He, Dongfang Yang, et al. 👉mdpi.com/2072-4292/11/9/1015… #remotesensing #deeplearning #semanticsegmentation #urbanplanning

画質をなるべくきれいなまま、ファイルの容量を削減したいですよね。画質調整は、#lambda の環境変数で指定することが可能です。視覚的な画質の変化は、#Structuralsimilarity を利用しています。 buff.ly/2wZeA32 #aws #s3