From e349d1b8b76cca8fd8e898268c202adacf7f9731 Mon Sep 17 00:00:00 2001 From: Scot Federman Date: Wed, 2 Apr 2014 13:25:50 -0700 Subject: [PATCH] Update Readme.txt --- Readme.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Readme.txt b/Readme.txt index cae8371..3e28eae 100644 --- a/Readme.txt +++ b/Readme.txt @@ -59,7 +59,7 @@ The steps to install SURPI on a machine are as follows: The content and creation of the SNAP databases is documented in the paper in the Reference Databases section, which is duplicated below: -A 3.1 gigabase (Gb) human nucleotide database (human DB) was constructed from a combination of human genomic DNA (GRCh37 / hg19), rRNA (RefSeq), mRNA (RefSeq), and mitochondrial RNA (RefSeq) sequences in NCBI as of March of 2012. The bacterial nucleotide, viral nucleotide, and viral protein databases used by SURPI in fast mode (bacterial DB, viral nucleotide DB, and viral protein DB, respectively) were also constructed from sequences in NCBI as of March of 2012. The 3 Gb bacterial DB was constructed from all bacterial RefSeq entries and consisted of 348,922 unique accessioned sequences, each with a minimum length of 100 bp. The 1.4 Gb viral nucleotide DB included 1,193,607 entries and was constructed by searching for all viral sequences in the 42 Gb National Center for Biotechnology Information (NCBI) nt collection using the query term �viridae[Organism]� in BioPython. The viral protein DB was similarly constructed by extracting viral sequences from the NCBI nr DB collection. Index tables for SNAP (v0.15) were generated with an empirically determined default seed size of 20 for the human DB and viral nucleotide DB, and seed size of 16 for the bacterial DB. Index tables for RAPSearch (v2.09) were generated from the viral protein DB using default parameters. +A 3.1 gigabase (Gb) human nucleotide database (human DB) was constructed from a combination of human genomic DNA (GRCh37 / hg19), rRNA (RefSeq), mRNA (RefSeq), and mitochondrial RNA (RefSeq) sequences in NCBI as of March of 2012. The bacterial nucleotide, viral nucleotide, and viral protein databases used by SURPI in fast mode (bacterial DB, viral nucleotide DB, and viral protein DB, respectively) were also constructed from sequences in NCBI as of March of 2012. The 3 Gb bacterial DB was constructed from all bacterial RefSeq entries and consisted of 348,922 unique accessioned sequences, each with a minimum length of 100 bp. The 1.4 Gb viral nucleotide DB included 1,193,607 entries and was constructed by searching for all viral sequences in the 42 Gb National Center for Biotechnology Information (NCBI) nt collection using the query term "viridae[Organism]" in BioPython. The viral protein DB was similarly constructed by extracting viral sequences from the NCBI nr DB collection. Index tables for SNAP (v0.15) were generated with an empirically determined default seed size of 20 for the human DB and viral nucleotide DB, and seed size of 16 for the bacterial DB. Index tables for RAPSearch (v2.09) were generated from the viral protein DB using default parameters. To generate the National Center for Biotechnology Information (NCBI) nucleotide (nt) collection (NCBI nt DB) used by SURPI in comprehensive mode, the complete 42 Gb nucleotide collection (nt) was downloaded from NCBI in January of 2013. This collection consists of a comprehensive archive of sequences from multiple sources, including GenBank, European Molecular Biology Laboratory (EMBL), DNA Data Bank of Japan (DDBJ), and Protein Data Bank (PDB), and is the richest collection of annotated microbial sequence data publicly available. As SNAP uses 32-bit offsets in the reference genome during hashing, the aligner restricts the size of the reference genome to an absolute maximum of 2^32 bases, or ~4.2 Gb. Thus, the 42 Gb NCBI nt collection was first split into 29 sub-databases, each approximately 1.5 Gb in size. Each sub-database was then indexed separately by SNAP at default parameters with a seed size of 20. This generated 29 SNAP indexed databases, each approximately 27 GB in size, with the aggregate of all 29 databases referred to as the NCBI nt DB.