PredictProtein (PP) is an internet service for sequence analysis and prediction of protein structure and function. Users submit protein sequences or alignments, and the server returns multiple sequence alignments, PROSITE motifs, low-complexity regions, domain assignments, nuclear localization signals, and predictions of secondary structure, solvent accessibility, globular regions, transmembrane helices, coiled-coil regions, structural switch regions, and disulfide bonds. Fold recognition by prediction-based threading is available upon request. Users can submit queries via email or interactively through a web interface. PP has been providing protein structure prediction services for over a decade. It was the first internet server for protein structure prediction and remains the most widely used public server for structure prediction. Over one million requests have been handled from users in 95 countries. PP's web pages are mirrored in 17 countries on four continents. The server uses a variety of methods to predict protein structure and function, including database searches, structure prediction methods, and domain assignments. These methods include BLAST, PSI-BLAST, MaxHom, PHD, PROF, COILS, CYSPRED, ASP, and others. PP aims to provide a comprehensive and easy-to-interpret set of predictions for a protein in one image. It incorporates a hierarchy of thresholds to simplify output interpretation. The server uses a dynamic programming method for sequence alignment and has a unique approach to PSI-BLAST searches and sequence family inclusion. PP's predictions include secondary structure, solvent accessibility, membrane helices, coiled-coil regions, and disulfide bonds. The server also provides information on the expected accuracy of structure predictions. PP supports various input formats, including one-letter residue sequences, FASTA, PIR, and SWISS-PROT. Users can submit their own alignments for improved prediction accuracy. The server also allows for advanced prediction and job options, including toggling methods and specifying thresholds. PP's output can be in various formats, including HTML, BLAST, HSSP, HSSP profiles, MSF, SAF, and FASTA. The server's performance is evaluated in terms of accuracy and reliability, with a focus on providing high-quality predictions. PP has received support from various funding sources and has been maintained by a team of researchers and developers.PredictProtein (PP) is an internet service for sequence analysis and prediction of protein structure and function. Users submit protein sequences or alignments, and the server returns multiple sequence alignments, PROSITE motifs, low-complexity regions, domain assignments, nuclear localization signals, and predictions of secondary structure, solvent accessibility, globular regions, transmembrane helices, coiled-coil regions, structural switch regions, and disulfide bonds. Fold recognition by prediction-based threading is available upon request. Users can submit queries via email or interactively through a web interface. PP has been providing protein structure prediction services for over a decade. It was the first internet server for protein structure prediction and remains the most widely used public server for structure prediction. Over one million requests have been handled from users in 95 countries. PP's web pages are mirrored in 17 countries on four continents. The server uses a variety of methods to predict protein structure and function, including database searches, structure prediction methods, and domain assignments. These methods include BLAST, PSI-BLAST, MaxHom, PHD, PROF, COILS, CYSPRED, ASP, and others. PP aims to provide a comprehensive and easy-to-interpret set of predictions for a protein in one image. It incorporates a hierarchy of thresholds to simplify output interpretation. The server uses a dynamic programming method for sequence alignment and has a unique approach to PSI-BLAST searches and sequence family inclusion. PP's predictions include secondary structure, solvent accessibility, membrane helices, coiled-coil regions, and disulfide bonds. The server also provides information on the expected accuracy of structure predictions. PP supports various input formats, including one-letter residue sequences, FASTA, PIR, and SWISS-PROT. Users can submit their own alignments for improved prediction accuracy. The server also allows for advanced prediction and job options, including toggling methods and specifying thresholds. PP's output can be in various formats, including HTML, BLAST, HSSP, HSSP profiles, MSF, SAF, and FASTA. The server's performance is evaluated in terms of accuracy and reliability, with a focus on providing high-quality predictions. PP has received support from various funding sources and has been maintained by a team of researchers and developers.