dorsal/arxiv
View SchemaGlobal statistical analysis of the protein homology network
| Authors | Concetta Miccio, Thomas Rattei |
|---|---|
| Categories | |
| ArXiv ID | q-bio/0703053 |
| URL | https://arxiv.org/abs/q-bio/0703053 |
Abstract
The similarity between protein sequences is a directly and easly computed quantity from which to deduce information about their evolutionary distance and to detect homologous proteins. The SIMAP database -- Similarity Matrix of Proteins -- provides a pre-computed similarity matrix covering the similarity space formed by about all publicly available amino acid sequences from public databases and completely sequenced genomes. From SIMAP we construct the protein homology network, where the proteins are the nodes and the links represent homology relationships. With more than 5 million nodes and about 70 10^9 edges it is the greatest protein homology network ever been builded. We describe the basic features and we perform a global statistical analysis of the network. Starting from the Smith-Waterman similarity score, we define for each edge a weight w to measure the similarity distance between two nodes. Keeping only edges with a weigth greater than a minimal w_min, and by varying w_min we build a family of networks with different degree of similarity. We investigate the distribution of connected components (clusters) of the networks at different w_min and in particular we find a behaviour similar to a phase transition guided by the formation of a giant component. Moreover we study selected sequence features and protein domains of protein pairs that connect different clusters in the networks at different level of similarity. We observed specific, non-random distributions of the protein features and domains for proteins connecting clusters at certain weight intervals.
{
"annotation_id": "2bff8318-1cb4-423c-b72a-23288cf25afa",
"date_created": "2026-03-02T18:01:34.806000Z",
"date_modified": "2026-03-02T18:01:34.806000Z",
"file_hash": "f8e7e887b48fb5d98e6c0da79b1cdeb5465ddbe992cc7401855c89159fc6953c",
"private": false,
"record": {
"abstract": "The similarity between protein sequences is a directly and easly computed\nquantity from which to deduce information about their evolutionary distance and\nto detect homologous proteins. The SIMAP database -- Similarity Matrix of\nProteins -- provides a pre-computed similarity matrix covering the similarity\nspace formed by about all publicly available amino acid sequences from public\ndatabases and completely sequenced genomes. From SIMAP we construct the protein\nhomology network, where the proteins are the nodes and the links represent\nhomology relationships. With more than 5 million nodes and about 70 10^9 edges\nit is the greatest protein homology network ever been builded. We describe the\nbasic features and we perform a global statistical analysis of the network.\nStarting from the Smith-Waterman similarity score, we define for each edge a\nweight w to measure the similarity distance between two nodes. Keeping only\nedges with a weigth greater than a minimal w_min, and by varying w_min we build\na family of networks with different degree of similarity. We investigate the\ndistribution of connected components (clusters) of the networks at different\nw_min and in particular we find a behaviour similar to a phase transition\nguided by the formation of a giant component. Moreover we study selected\nsequence features and protein domains of protein pairs that connect different\nclusters in the networks at different level of similarity. We observed\nspecific, non-random distributions of the protein features and domains for\nproteins connecting clusters at certain weight intervals.",
"arxiv_id": "q-bio/0703053",
"authors": [
"Concetta Miccio",
"Thomas Rattei"
],
"categories": [
"q-bio.QM",
"q-bio.MN"
],
"title": "Global statistical analysis of the protein homology network",
"url": "https://arxiv.org/abs/q-bio/0703053"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "e6f568b1-913a-42f1-a6c6-089c15d8701e",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}