estlibcount.cpp File Reference

#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <set>
#include <cstdlib>
#include <cstring>
#include <cassert>
#include "strformat.h"
#include <boost/math/distributions/chi_squared.hpp>

Classes
class	lesschptr
Functions
void	readIdList (set< string > &ids, const string &file)
void	readIdList (set< const char *, lesschptr > &ids, const string &file)
void	readIdFromFasta (set< const char *, lesschptr > &ids, const string &file)
vector< int >	processModels (const string &modfile, const vector< set< const char *, lesschptr > > &estlibs, const vector< string > &libfiles, vector< int * > &result, ostream &log)
int	hitCount (const string &ids, const vector< set< const char *, lesschptr > > &estlib, vector< set< const char *, lesschptr > > &mapped, int *&counts)
int	countESTIDs (const string &modfile)
void	usage ()
void	releaseMemory (vector< set< const char *, lesschptr > > &estlibs)
void	releaseResult (vector< int * > &res)
void	ChisquareAndRatio (vector< int * > &counts, vector< int > &overallcount, ostream &ous, ostream &RAE, ostream &log)
void	outputHeader (ostream &ous, const vector< string > &libfiles)
string	nameraefile (const string &countfile)
int	main (int argc, char *argv[])

---

Function Documentation

void ChisquareAndRatio	(	vector< int * > &	counts,
		vector< int > &	overallcount,
		ostream &	ous,
		ostream &	RAE,
		ostream &	log
	)

References N.

Referenced by main().

int countESTIDs

(

const string &

modfile

)

since one est can be assembled into multiple models, I have to produce a unique list from the input models

References ifstream().

Referenced by main().

int hitCount	(	const string &	ids,
		const vector< set< const char *, lesschptr > > &	estlib,
		vector< set< const char *, lesschptr > > &	mapped,
		int *&	counts
	)

Parameters:

counts

store the number of hits in each library

This is the pointer version and should be more efficient in both space and CPU speed. pointer version faster. This is used for production.

Parameters:

	ids	is the id string from each combest model result. Ids are separated with comma "FYT9WAP01BTJGL,FYT9WAP01CKLAO,FYT9WAP01DZQZM". This is the input. This function will simply convert this list into a count for each input library as given by estlib and mapped.
	estlib	a vector of set of estids from each library. This is the other input for this function. This contain the ESTids from each input library. Not all of the ESTs may be mapped to genome and so will not be included into any models.
	mapped	this is a result accumulator that contains the exact ESTids that are mapped from each EST library. There is a one-to-one library conrespondence between this and estlib. They should have the same size unless for some reason, none of the EST from one input estlib maps.
	counts	is the output result for this particular model. This result vector has the dimension of one extra than estlib the last element is used for holding modelid. So you have 10 EST ids after this function call, counts could be |3|0|5|2|formodid| with four input library.

Returns:

the total number of EST mapped from the provided est library. Since the model construction may use EST sequence from libraries other than the library of interest, this number could be zero even a model may have many ESTs.

This funciton will not write result to output stream.

References find(), and split().

Referenced by processModels().

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

References ChisquareAndRatio(), countESTIDs(), nameraefile(), outputHeader(), processModels(), readIdFromFasta(), readIdList(), releaseMemory(), releaseResult(), and usage.

string nameraefile

(

const string &

countfile

)

Referenced by main().

void outputHeader	(	ostream &	ous,
		const vector< string > &	libfiles
	)

References find().

Referenced by main().

vector< int > processModels	(	const string &	modfile,
		const vector< set< const char *, lesschptr > > &	estlibs,
		const vector< string > &	libfiles,
		vector< int * > &	result,
		ostream &	log
	)

pointer version for production. This is more efficient. This is the main function for doing the main job.

Parameters:

	modfile	the combest output file name. This funciton will only use the ESTids column.
	estlibs	is the input est library ESTids preloaded.
	libfiles	the library file names contain lists of ESTids. This is only used for output headers for the result file.
	result	is a vector of mapping result for each input model. The last column is modelid.

Returns:

the overal mapping count in each input EST library

References hitCount(), and ifstream().

Referenced by main().

void readIdFromFasta	(	set< const char *, lesschptr > &	ids,
		const string &	file
	)

References ifstream().

Referenced by main().

void readIdList	(	set< const char *, lesschptr > &	ids,
		const string &	file
	)

This is the memory efficient version, and faster Given a files of ESTIds. The format of the file is one or more ESTids separated by one or more space (tab or space). This function will store the result into the set.

Parameters:

	file	is usually named after the library that contains all ESTids from one condition.
	ids	container to hold the reusult. The result pointer memory needs to be freed with delete.

References ifstream().

void readIdList	(	set< string > &	ids,
		const string &	file
	)

Older version using set<string>, takes too much memory replaced with pointer version set<char*, lesschptr> this function will get rid of redundant entires. id file format 1. >id one or more per line 2. id1 id2 ... 3. fasta file

References ifstream().

Referenced by main().

void releaseMemory

(

vector< set< const char *, lesschptr > > &

estlibs

)

Referenced by clustergene(), groupModel(), and main().

void releaseResult

(

vector< int * > &

res

)

Referenced by main().

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.