getSRA: Fulltext search SRA meta data using SQLite fts3 module

In SRAdb: A compilation of metadata from NCBI SRA and tools

Description

This function does Fulltext search on any SRA fields in any SRA data types with Fulltext capacity in the SQLite and returns SRA records

Usage

getSRA(search_terms, out_types=c('sra','submission','study','experiment','sample','run'), sra_con, acc_only=FALSE)

Arguments

search_terms	Free text search terms constructed according to SQLite query syntax defined here: http://www.sqlite.org/fts3.html#section_1_3
out_types	Character vector of the following SRA data types: 'sra','submission','study','sample','experiment','run'. Note: if 'sra' is within out_types, the out_types will be set to c('submission','study','sample','experiment').
sra_con	Connection to the SRAmetadb SQLite database
acc_only	logical, if TRUE, the function will return SRA accession for each out_types

Details

Queries performed by this function could be Phrase queries, e.g. '"lin* app*"', or NEAR queries, e.g. '"ACID compliant" NEAR/2 sqlite', or with the Enhanced Query Syntax. Check Full Text Search section on the SQLite site for details. if 'acc_only=TRUE', a data.frame containing only SRA accessions will be returned, which can be used as input for sraGraph.

Value

A data.frame containing all returned SRA records with fields defined by out_types.

If acc_only=FALSE, a data.frame of matched accessions of out_types will be returned.

Author(s)

Jack Zhu <[e-mail]>

References

http://www.sqlite.org/

See Also

sraConvert

Examples

## Using the SRAmetadb demo database

	library(SRAdb)
	sra_dbname <- file.path(system.file('extdata', package='SRAdb'), 'SRAmetadb_demo.sqlite')	
	sra_con <- dbConnect(dbDriver("SQLite"), sra_dbname)
  
	## Fulltext search SRA meta data using SQLite fts3 module:
	# find all records with words of 'breast' and 'cancer' in a filed and there could be one to many words between 'breast' and 'cancer':   
	rs <- getSRA (search_terms ='breast cancer', out_types=c('run','study'), sra_con=sra_con)
 
	# find all records with exact phrase of 'breast cancer' in a filed:   
	rs <- getSRA (search_terms ='"breast cancer"', out_types=c('run','study'), sra_con=sra_con)
 
	# find records with words beginning with 'braes' and 'can', and the distance between them is equal or less than two words:
	rs <- getSRA (search_terms ='breas* NEAR/2 can*', out_types=c('run','study'), sra_con=sra_con)

	# the same as above except that only one space between the two words  
	rs <- getSRA (search_terms ='"breas* can*"', out_types=c('study'), sra_con=sra_con) 
 
	# find records with 'MCF7' or 'MCF-7' - adding double quote to avoid the SQLite to break down 'MCF-7' to 'MCF' and '7':
	rs <- getSRA (search_terms ='MCF7 OR "MCF-7"', out_types=c('sample'), sra_con=sra_con) 

	# the same as above, but only search the field of 'study_title':
	rs <- getSRA (search_terms ='study_title: brea* can*', out_types=c('run','study'), sra_con=sra_con)  

	# the same as above, but only search the field of 'study_title' and return only accessions:
	rs <- getSRA (search_terms ='study_title: brea* can*', out_types=c('run','study'), sra_con=sra_con, acc_only=TRUE) 

## The actual SRAmetadb sqlite database can be downloaded using function: getSRAdbFile. Warning: the actual SRAmetadb sqlite database is pretty large (> 35GB as of May, 2018) after uncompression. So, downloading and uncompressing of the actual SRAmetadb sqlite could take quite a few minutes depending on your network bandwidth. Direct links for downloading the SRAmetadb sqlite database:  https://s3.amazonaws.com/starbuck1/sradb/SRAmetadb.sqlite.gz https://gbnci-abcc.ncifcrf.gov/backup/SRAmetadb.sqlite.gz

SRAdb documentation built on Nov. 8, 2020, 6:49 p.m.