dbaln_fungal.cpp File Reference

#include "mysql++.h"
#include <string>
#include "bioseq.h"
#include "dynaln.h"
#include <math.h>
#include <fstream>
#include "dbinfo.h"
#include <map>
#include "strformat.h"
#include "dbalnhelper.h"

Classes
struct	ProteinTable
Defines
#define	MYSQLPLUS2
Functions
void	getPrt (mysqlpp::Query &q, const string &id, string &prt)
void	alignProtein (const string &alnfile, const string &alntabfile, const StoreQueryResult &res, Connection &conn1, Connection &conn2, const ProteinTable &pt1, const ProteinTable &pt2)
void	alignDNA (const string &alnfile, const string &alntabfile, const string &pairtab, mysqlpp::Connection &conn, mysqlpp::Connection &conn1, mysqlpp::Connection &conn2)
void	bufferGenomic (mysqlpp::Connection &conn, const string &sqlstr, map< string, DNA > &id2seq)
void	usage ()
void	bufferProtein (map< int, string > &store, Connection &conn, const ProteinTable &pt)
int	main (int argc, char *argv[])
Variables
string	user
string	passwd

---

Define Documentation

#define MYSQLPLUS2

---

Function Documentation

void alignDNA	(	const string &	alnfile,
		const string &	alntabfile,
		const string &	pairtab,
		mysqlpp::Connection &	conn,
		mysqlpp::Connection &	conn1,
		mysqlpp::Connection &	conn2
	)

void alignProtein	(	const string &	alnfile,
		const string &	alntabfile,
		const StoreQueryResult &	res,
		Connection &	conn1,
		Connection &	conn2,
		const ProteinTable &	pt1,
		const ProteinTable &	pt2
	)

References bufferProtein(), Dynaln::headers(), itos(), length, Dynaln::printAlign(), Dynaln::runlocal(), and Dynaln::toDelimitedString().

void bufferGenomic	(	mysqlpp::Connection &	conn,
		const string &	sqlstr,
		map< string, DNA > &	id2seq
	)

void bufferProtein	(	map< int, string > &	store,
		Connection &	conn,
		const ProteinTable &	pt
	)

taken the protein sequence from the JGI protein table this table has column (proteinId, seq)

The two columns can be given as two optional arguments for NIBI proteinseq table it has (gi, sequence) for Broad proteinseq columns (id, seq)

References ProteinTable::idcol, ProteinTable::name, ProteinTable::seqcol, and string().

Referenced by alignProtein(), and bufferAllProteins().

void getPrt	(	mysqlpp::Query &	q,
		const string &	id,
		string &	prt
	)

the tab must have id and seq column

int main	(	int	argc,
		char *	argv[]
	)

References alignProtein(), MysqlDBInfo::getAuthenInfo(), MysqlDBInfo::getPassword(), MysqlDBInfo::getUser(), Dynaln::headers(), passwd, ProteinTable::set(), usage, and user.

void usage

(

)

given gmap summary format, this program converts it into combest archive format (*.car)

It can be used a pipe, or given specific file names.

this is a helper program to count distinct estids in the ESTId column of the combest result. So it is the actual number of ESTs mapped. This could be lower for deeply covered genome because of the coverage depth-dependent filtering.

given gmap summary format, this program converts it into combest archive format (*.car)

It can be used a pipe, or given specific file names.

this is a helper program to count distinct estids in the ESTId column of the combest result. So it is the actual number of ESTs mapped. This could be lower for deeply covered genome because of the coverage depth-dependent filtering.

---

Variable Documentation

string passwd

string user

use three connections, the first connection provide the alignment pairs. the second and the third provided the raw sequences All three connections can be the same. each connection provide a table.

format I am adding a format for bootstrap where a vector of 0,1,2 are output that stands for mismatch, similar, and match respectively.