Low-Complexity Proteins Ortholog Database

This dataset contains the core relational data and multiple sequence alignments (MSAs) for the Low-Complexity Proteins Ortholog Database. It provides a comprehensive, curated reference of low-complexity regions (LCRs) identified in human proteins mapped to their vertebrate orthologs (e.g., Mouse, Chimp, Macaque, Cow, Rat, Pig, Dog, Opossum, Chicken, Xenopus, and Zebrafish). The dataset is designed to improve confidence in ortholog matches through quality-focused curation, filtering out noise and ghost alignments using entropy, species mapping, and anchor-flank local sequence alignments. Contents of the Dataset: lcr_ortholog_tables_v1.zip: Contains the relational database structure in CSV format, ready for PostgreSQL ingestion. protein.csv: Full-length protein records and Uniprot accessions. lcr_seq.csv: Individual LCR sequences ("atoms") extracted using SEG (default and strict parameters). lcr_family.csv: Ortholog groupings. phylogeny_map.csv: The anchor mapping evidence linking LCRs to their ortholog families, including percent identity, coverage, and E-values. lcr_biophys.csv: Biophysical features per LCR (e.g., fraction of charged residues, net charge per residue, mean hydropathy) calculated using localCIDER metrics. lcr_function_class.csv & lcr_to_class.csv: Functional class label lookups and associations (e.g., Acidic Activation Domains). lcr_ortholog_fastas_v1.zip: Contains the grouped FASTA files for the anchored MSAs. anchor_fastas_default/: Generated FASTA groups using default LCR parameters. anchor_fastas_strict/: Generated FASTA groups using strict LCR parameters. Usage:This data is intended to be used in conjunction with the pipeline tools and schema available in the project GitHub repository. After downloading and extracting these files into the data directory of the repository, the provided schema scripts can be executed to fully instantiate the local database.