main.cpp File Reference

#include <array>
#include <cstring>
#include <iostream>
#include <chrono>
#include <iomanip>
#include <fstream>
#include <vector>
#include <algorithm>
#include <stdexcept>
#include <omp.h>
#include <immintrin.h>
#include <mpi.h>
#include <sstream>
#include "EDNAFULL.h"
#include "EBLOSUM62.h"
#include <unordered_map>
#include <filesystem>
#include <climits>

Data Structures
struct	AffineDPScores
struct	Loc

Macros
#define	RESET   "\033[0m"
	ANSI escape code: reset color.
#define	GREEN   "\033[32m"
	ANSI escape code: green.
#define	RED   "\033[31m"
	ANSI escape code: red.
#define	CYAN   "\033[36m"
	ANSI escape code: cyan.

Typedefs
using	ScoreFn = int(*)(char, char)

Enumerations
enum	ScoreMode { MODE_DNA , MODE_PROTEIN }

Functions
int	edna_score (char x, char y)
	Lookup the score between two characters based on the selected mode.
int	blosum62_score (char x, char y)
	Lookup the score between two characters based on the selected mode.
int	score (char x, char y, ScoreMode mode)
	Lookup the score between two characters based on the selected mode.
void	showProgressBar (int progress, int total)
	Show a progress bar in the console.
std::string	getAccession (const std::string &header, ScoreMode mode)
	Extract the accession number from a FASTA header line.
std::string	getGeneSymbol (const std::string &header, ScoreMode mode)
	Extract the gene symbol from a FASTA header.
void	processFasta (const string &filename, string &header, string &sequence)
	Read the first record from a FASTA file.
void	savePlainAlignment (const std::string &header1, const std::string &header2, const std::string &aligned1, const std::string &aligned2, std::ostream &os)
	Write two aligned sequences in FASTA format.
void	saveLCS (const std::string &id, const std::string &lcs_str, std::ostream &os)
	Write the LCS (Longest Common Subsequence) in FASTA format.
void	printColoredAlignment (const std::string &seq1_aln, const std::string &seq2_aln, std::ostream &os=std::cout)
	Print two aligned sequences with color coding.
void	writeRawDPMatrix (const std::vector< std::vector< int > > &dp, const std::string &filename)
	Write the DP matrix to a file.
void	writeDPMatrix (const std::vector< std::vector< int > > &dp, const std::string &filename)
	Write the DP matrix to a binary file.
void	writeRawCharMatrix (const std::vector< std::vector< char > > &mat, const std::string &filename)
	Write a character matrix to a file.
void	writeCharMatrix (const std::vector< std::vector< char > > &mat, const std::string &filename)
	Write a character matrix to a binary file.
void	initAffineDP (int n, vector< int > &prev_row, vector< int > &prev_gapX, vector< int > &prev_gapY, bool isGlobal)
	Initialize the DP structures for affine gap scoring.
void	computeAffineDPRow (int i, const string &x, const string &y, vector< int > &prev_row, vector< int > &prev_gapX, vector< int > &prev_gapY, vector< int > &curr_row, vector< int > &curr_gapX, vector< int > &curr_gapY, vector< char > &curr_trace_row, ScoreFn score_fn)
	Compute a single row of the affine DP matrix.
void	globalalign (const string &x, const string &y, const string &header1, const string &header2, const std::string &outdir, ScoreMode mode, ScoreFn score_fn)
	Perform global sequence alignment using the Needleman-Wunsch algorithm.
AffineDPScores	compute_local_affine_cell (int s_diag_prev, int e_diag_prev, int f_diag_prev, int s_left, int e_left, int s_up, int f_up, char char_x, char char_y, ScoreFn score_fn)
	Compute a single cell in the local affine DP matrix.
void	localalign (const std::string &x, const std::string &y, const std::string &header1, const std::string &header2, const std::string &outdir, ScoreMode mode, ScoreFn score_fn)
	Perform local sequence alignment using the Smith-Waterman algorithm.
void	lcs (const string &x, const string &y, const string &header1, const string &header2, const std::string &outdir, ScoreMode mode)
	Compute the Longest Common Subsequence (LCS) of two sequences.
int	main (int argc, char **argv)
	Main function to run the MPI-based sequence alignment tool.

Variables
bool	verbose = false
bool	binary = false
bool	txt = false
double	GAP_OPEN = -5.0
	Gap penalty.
double	GAP_EXTEND = -1.0

Macro Definition Documentation

CYAN

#define CYAN   "\033[36m"

ANSI escape code: cyan.

GREEN

#define GREEN   "\033[32m"

ANSI escape code: green.

RED

#define RED   "\033[31m"

ANSI escape code: red.

RESET

#define RESET   "\033[0m"

ANSI escape code: reset color.

Typedef Documentation

ScoreFn

using ScoreFn = int(*)(char,char)

Enumeration Type Documentation

ScoreMode

enum ScoreMode

Enumerator
MODE_DNA
MODE_PROTEIN

Function Documentation

blosum62_score()

int blosum62_score ( char  x, char  y  )

inline

Lookup the score between two characters based on the selected mode.

Parameters

x	First character (base or amino acid).
y	Second character (base or amino acid).

Returns

Score based on the selected scoring matrix.

compute_local_affine_cell()

AffineDPScores compute_local_affine_cell	(	int	s_diag_prev,
		int	e_diag_prev,
		int	f_diag_prev,
		int	s_left,
		int	e_left,
		int	s_up,
		int	f_up,
		char	char_x,
		char	char_y,
		ScoreFn	score_fn
	)

Compute a single cell in the local affine DP matrix.

This function computes the scores for a single cell in the local alignment DP matrix using affine gap penalties. It returns an AffineDPScores struct containing the scores and the pointer for traceback.

Parameters

s_diag_prev	S value from the diagonal cell (i-1, j-1)
e_diag_prev	E value from the diagonal cell (i-1, j-1)
f_diag_prev	F value from the diagonal cell (i-1, j-1)
s_left	S value from the left cell (i, j-1)
e_left	E value from the left cell (i, j-1)
s_up	S value from the upper cell (i-1, j)
f_up	F value from the upper cell (i-1, j)
char_x	Character from sequence X at position i
char_y	Character from sequence Y at position j
score_fn	Scoring function for match/mismatch

Returns

AffineDPScores containing computed scores and pointer

computeAffineDPRow()

void computeAffineDPRow	(	int	i,
		const string &	x,
		const string &	y,
		vector< int > &	prev_row,
		vector< int > &	prev_gapX,
		vector< int > &	prev_gapY,
		vector< int > &	curr_row,
		vector< int > &	curr_gapX,
		vector< int > &	curr_gapY,
		vector< char > &	curr_trace_row,
		ScoreFn	score_fn
	)

Compute a single row of the affine DP matrix.

This function computes the i-th row of the DP matrix for affine gap scoring, including match/mismatch scores and gap scores in both directions.

Parameters

i	The current row index (1-based).
x	The first sequence (string).
y	The second sequence (string).
prev_row	Previous row S_{i-1}.
prev_gapX	Previous gap row F_{i-1} (gap in Y).
prev_gapY	Previous gap row E_{i-1} (gap in X).
curr_row	Current row S_i to be filled.
curr_gapX	Current gap row F_i to be filled.
curr_gapY	Current gap row E_i to be filled.
curr_trace_row	Pointer/Trace for S_i to be filled.
score_fn	Scoring function for match/mismatch.

edna_score()

int edna_score ( char  x, char  y  )

inline

Lookup the score between two characters based on the selected mode.

Parameters

x	First character (base or amino acid).
y	Second character (base or amino acid).

Returns

Score based on the selected scoring matrix.

getAccession()

std::string getAccession	(	const std::string &	header,
		ScoreMode	mode
	)

Extract the accession number from a FASTA header line.

The accession number is the first word in the header line, which is expected to start with a >.

Parameters

header

The header line from a FASTA file.

Returns

The accession number (first word).

getGeneSymbol()

std::string getGeneSymbol	(	const std::string &	header,
		ScoreMode	mode
	)

Extract the gene symbol from a FASTA header.

For DNA headers, looks for the first pair of parentheses “(GENE)” and returns GENE. For protein headers, takes the part after the second ‘|’ up to the first underscore.

Parameters

header	The FASTA header line (without the leading '>').
mode	MODE_DNA or MODE_PROTEIN.

Returns

The gene symbol, or an empty string on failure.

globalalign()

void globalalign	(	const string &	x,
		const string &	y,
		const string &	header1,
		const string &	header2,
		const std::string &	outdir,
		ScoreMode	mode,
		ScoreFn	score_fn
	)

Perform global sequence alignment using the Needleman-Wunsch algorithm.

This function computes the global alignment of two sequences using affine gap scoring. It initializes the DP matrix, fills it row by row, and performs traceback to get the aligned sequences.

Parameters

x	The first sequence (string).
y	The second sequence (string).
header1	Header for the first sequence.
header2	Header for the second sequence.
outdir	Output directory for results.
mode	Scoring mode (MODE_DNA or MODE_PROTEIN).
score_fn	Scoring function for match/mismatch.

initAffineDP()

void initAffineDP	(	int	n,
		vector< int > &	prev_row,
		vector< int > &	prev_gapX,
		vector< int > &	prev_gapY,
		bool	isGlobal
	)

Initialize the DP structures for affine gap scoring.

Parameters

n	Length of the second sequence (Y).
prev_row	Row i-1, match/mismatch scores.
prev_gapX	Row i-1, gap scores in X.
prev_gapY	Row i-1, gap scores in Y.
isGlobal	Whether this is a global alignment (affects initialization).

lcs()

void lcs	(	const string &	x,
		const string &	y,
		const string &	header1,
		const string &	header2,
		const std::string &	outdir,
		ScoreMode	mode
	)

Compute the Longest Common Subsequence (LCS) of two sequences.

This function computes the LCS using dynamic programming and saves the results.

Parameters

x	The first sequence (string).
y	The second sequence (string).
header1	Header for the first sequence.
header2	Header for the second sequence.
outdir	Output directory for results.
mode	Scoring mode (MODE_DNA or MODE_PROTEIN).

localalign()

void localalign	(	const std::string &	x,
		const std::string &	y,
		const std::string &	header1,
		const std::string &	header2,
		const std::string &	outdir,
		ScoreMode	mode,
		ScoreFn	score_fn
	)

Perform local sequence alignment using the Smith-Waterman algorithm.

This function computes the local alignment of two sequences using affine gap scoring. It initializes the DP matrix, fills it row by row, and performs traceback to get the aligned sequences.

Parameters

x	The first sequence (string).
y	The second sequence (string).
header1	Header for the first sequence.
header2	Header for the second sequence.
outdir	Output directory for results.
mode	Scoring mode (MODE_DNA or MODE_PROTEIN).
score_fn	Scoring function for match/mismatch.

main()

int main	(	int	argc,
		char **	argv
	)

Main function to run the MPI-based sequence alignment tool.

This function initializes MPI, parses command line arguments, reads input files, and performs sequence alignments based on user choices.

Parameters

argc	Number of command line arguments.
argv	Array of command line arguments.

Returns

Exit status code.

printColoredAlignment()

void printColoredAlignment	(	const std::string &	seq1_aln,
		const std::string &	seq2_aln,
		std::ostream &	os = std::cout
	)

Print two aligned sequences with color coding.

Matches are printed in green, mismatches in cyan, and gaps in red. The alignment is printed in blocks of LINE_WIDTH characters.

Parameters

seq1_aln	Aligned sequence 1 (with gaps).
seq2_aln	Aligned sequence 2 (with gaps).
os	Output stream (default is std::cout).

processFasta()

void processFasta	(	const string &	filename,
		string &	header,
		string &	sequence
	)

Read the first record from a FASTA file.

Extracts the very first header (minus the leading >), and concatenates all subsequent lines into a single sequence string.

Parameters

	filename	Path to the FASTA file.
[out]	header	On return, the header line without leading >.
[out]	sequence	On return, the full sequence (no newlines).

Exceptions

std::runtime_error

if the file cannot be opened.

saveLCS()

void saveLCS	(	const std::string &	id,
		const std::string &	lcs_str,
		std::ostream &	os
	)

Write the LCS (Longest Common Subsequence) in FASTA format.

Parameters

id	Identifier for the LCS (no leading '>').
lcs_str	The LCS string.
os	Output stream (e.g. ofstream).

savePlainAlignment()

void savePlainAlignment	(	const std::string &	header1,
		const std::string &	header2,
		const std::string &	aligned1,
		const std::string &	aligned2,
		std::ostream &	os
	)

Write two aligned sequences in FASTA format.

Parameters

header1	Identifier for sequence 1 (no leading '>').
header2	Identifier for sequence 2 (no leading '>').
aligned1	Aligned sequence 1 (with gaps).
aligned2	Aligned sequence 2 (with gaps).
os	Output stream (e.g. ofstream).

score()

int score ( char  x, char  y, ScoreMode  mode  )

inline

Lookup the score between two characters based on the selected mode.

Parameters

x	First character (base or amino acid).
y	Second character (base or amino acid).
mode	Scoring mode (MODE_DNA or MODE_PROTEIN).

Returns

Score based on the selected scoring matrix.

showProgressBar()

void showProgressBar	(	int	progress,
		int	total
	)

Show a progress bar in the console.

*

Parameters

progress	Current progress (0 to total).
total	Total number of steps.

writeCharMatrix()

void writeCharMatrix	(	const std::vector< std::vector< char > > &	mat,
		const std::string &	filename
	)

Write a character matrix to a binary file.

Parameters

mat	The character matrix (2D vector).
filename	The output filename.

writeDPMatrix()

void writeDPMatrix	(	const std::vector< std::vector< int > > &	dp,
		const std::string &	filename
	)

Write the DP matrix to a binary file.

Parameters

dp	The DP matrix (2D vector).
filename	The output filename.

writeRawCharMatrix()

void writeRawCharMatrix	(	const std::vector< std::vector< char > > &	mat,
		const std::string &	filename
	)

Write a character matrix to a file.

Parameters

mat	The character matrix (2D vector).
filename	The output filename.

writeRawDPMatrix()

void writeRawDPMatrix	(	const std::vector< std::vector< int > > &	dp,
		const std::string &	filename
	)

Write the DP matrix to a file.

Parameters

dp	The DP matrix (2D vector).
filename	The output filename.

Variable Documentation

binary

bool binary = false

GAP_EXTEND

double GAP_EXTEND = -1.0

GAP_OPEN

double GAP_OPEN = -5.0

Gap penalty.

txt

bool txt = false

verbose

bool verbose = false