1. Gene Cluster Identification
LoGIC (Gene-cluster Integrative Catalog), a local online database that catalogs 20,306 local homologous gene clusters in 30 representative species.
1.1 Gene Cluster Definition
Gene clusters in LoGIC are defined as local groups of neighboring genes which have both shared InterPro domain and protein sequence similarity.
1.2 Gene Cluster Identification
For each species, genes were sorted first by the InterPro domain ID of their encoded proteins and then by their genomic locations. Distances between adjacent genes encoding common InterPro ID-containing proteins were calculated, and a two-component Gaussian Mixture Model (GMM) was applied to estimate the distance threshold for selecting gene clusters within each species. This threshold was then used as the MeanShift bandwidth to identify candidate gene clusters. Adjacent gene clusters with at least one shared common InterPro domain and genomic distance less than half of the MeanShift bandwidth were merged together. To ensure their homologous relationships, pairwise BLASTP was applied to the protein sequences among members of each cluster. The final members within each gene cluster were selected using a dual-evidence graph method, which considers both sequence similarity and shared InterPro domains (Figure 1.1).
2. Website Guidance
2.1 Overview
The Overview function provides a summary of species and gene clusters presented in LoGIC, as well as the overall statistics of gene clusters identified for each species, including the total number of gene clusters and the total number of clustered genes (Figure 2.1.1), the distribution of cluster lengths and the statistics of cluster sizes (Figure 2.1.2), together with the basic information on the top 20 largest gene clusters of each species (Figure 2.1.3).
2.2 Species
The Species module enables users to browse gene clusters identified in each species. A summary table for gene clusters in each species is provided (Figure 2.2.1), and users can invoke the detailed information of all gene cluster for each species (Figure 2.2.2 – Figure2.2.4) by clicking on the species name either at the species index or in the summary table.
2.3 Search
The Search module offers two search methods: Search by Gene and Search by Feature.
Search by Gene allows users to search for the cluster information of targeted gene by their names (Figure 2.3.1).
Search by Feature enables users to search and browse gene clusters with certain features, such as chromosome location, IPR family, cluster size, and GO annotation (Figure 2.3.2).
The search result page first displays the list of gene clusters matching the search criteria (Figure 2.3.3). By clicking on the name of each gene cluster, users can invoke a new webpage containing the general information and gene members of the cluster, a genome browser showing the positional relationship among cluster members and other neighboring genes, the multiple sequence alignment results of representative proteins encoded by each cluster member, and the Gene Ontology analysis of all gene member in the cluster (Figure 2.3.4). For analysis of human gene clusters, the corresponding Hi-C information and GTEx based expression will also be available (Figure 2.3.5).
2.4 Function
The Function module provides the GO term enrichment analysis results for all gene clusters of each species. Users can browse GO terms enriched by all gene clusters (Figure 2.4.1), or select certain GO terms according to the GO term hierarchical structure (Figure 2.4.2), or directly entering a specific GO term via the Input function (Figure 2.4.3).
In the GO Enrichment result page, the FDR score can be sorted to identify the most or least enriched GO terms (Figure 2.4.4). The 'is significant' checkbox enables users to only view gene clusters with functional enrichment (Figure 2.4.5). When clicking on the name of a cluster, a GO terms matrix related to all gene members within the cluster will be displayed on the right (Figure 2.4.5).
2.5 Evolution
The Evolution module displays the one-to-one orthologous relationships between genes, with humans as the reference species and other species as the target species. The conservation score reflects the degree of conservation of a cluster across different species (Figure 2.5.1). All columns in the conservation result table can be sorted to facilitate users to identify clusters meeting their needs. And a conservation matrix showing the conservation status of each gene member in other species will be invoked by clicking on any information of the selected cluster in the conservation table (Figure 2.5.1).
In the Cross-species Orthology function generates a Sankey diagram showing the pairwise conservation information of human genes of a selected cluster with their orthologs in other selected species (Figure 2.5.2), which is particularly useful for studying the expansion, reduction, splitting or aggregation of gene clusters across species.
Users can further compare the GO terms gene members between a human cluster with those in corresponding orthologous clusters in other species via Function Comparison (Figure 2.5.3).
