int
main
(
int argc,
char **argv
)
{
// Backtrace handler
signal(SIGSEGV, SIGSEGV_handler);
char *expr, *input;
// Unset flags (value -1).
int showdist_flag = -1;
int showpos_flag = -1;
int printline_flag = -1;
int matchonly_flag = -1;
int showline_flag = -1;
int count_flag = -1;
int invert_flag = -1;
int compact_flag = -1;
int dist_flag = -1;
int verbose_flag = -1;
int endline_flag = -1;
int prefix_flag = -1;
int best_flag = -1;
int nondna_flag = -1;
int memory_flag = -1;
int all_flag = -1;
// Unset options (value 'UNSET').
input = NULL;
if (argc == 1) {
say_version();
say_usage();
return EXIT_SUCCESS;
}
int c;
while (1) {
int option_index = 0;
static struct option long_options[] = {
{"positions", no_argument, 0, 'p'},
{"match-only", no_argument, 0, 'm'},
{"no-printline", no_argument, 0, 'n'},
{"print-dist", no_argument, 0, 'k'},
{"nondna", required_argument, 0, 'x'},
{"lines", no_argument, 0, 'l'},
{"count", no_argument, 0, 'c'},
{"invert", no_argument, 0, 'i'},
{"format-compact",no_argument, 0, 'f'},
{"verbose", no_argument, 0, 'z'},
{"version", no_argument, 0, 'v'},
{"help", no_argument, 0, 'h'},
{"end", no_argument, 0, 'e'},
{"prefix", no_argument, 0, 'r'},
{"best", no_argument, 0, 'b'},
{"all", no_argument, 0, 'a'},
{"memory", required_argument, 0, 'y'},
{"distance",required_argument, 0, 'd'},
{0, 0, 0, 0}
};
c = getopt_long(argc, argv, "apmnilczfvkherby:d:x:",
long_options, &option_index);
/* Detect the end of the options. */
if (c == -1) break;
switch (c) {
case 'd':
if (dist_flag < 0) {
int dist = atoi(optarg);
if (dist < 0) {
say_version();
fprintf(stderr, "error: distance must be a positive integer.\n");
say_help();
return EXIT_FAILURE;
}
dist_flag = atoi(optarg);
}
else {
say_version();
fprintf(stderr, "error: distance option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'y':
if (memory_flag < 0) {
int memory = atoi(optarg);
if (memory < 0) {
say_version();
fprintf(stderr, "error: memory limit must be a positive integer.\n");
say_help();
return EXIT_FAILURE;
}
memory_flag = atoi(optarg);
}
else {
say_version();
fprintf(stderr, "error: memory option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'v':
say_version();
return EXIT_SUCCESS;
case 'z':
if (verbose_flag < 0) {
verbose_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: verbose option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'r':
if (prefix_flag < 0) {
prefix_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: 'prefix' option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'a':
if (all_flag < 0) {
all_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: 'all' option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'x':
if (nondna_flag < 0) {
int nondna = atoi(optarg);
if (nondna < 0 || nondna > 2) {
say_version();
fprintf(stderr, "error: nondna value must be either 0, 1 or 2.\n");
say_help();
return EXIT_FAILURE;
}
nondna_flag = nondna;
}
else {
say_version();
fprintf(stderr, "error: 'nondna' option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'b':
if (best_flag < 0) {
best_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: 'best' option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'e':
if (endline_flag < 0) {
endline_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: line-end option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'p':
if (showpos_flag < 0) {
showpos_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: show-position option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'm':
if (matchonly_flag < 0) {
matchonly_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: match-only option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'n':
if (printline_flag < 0) {
printline_flag = 0;
}
else {
say_version();
fprintf(stderr, "error: no-printline option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'k':
if (showdist_flag < 0) {
showdist_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: show-distance option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'l':
if (showline_flag < 0) {
showline_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: show-line option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'c':
if (count_flag < 0) {
count_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: count option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'i':
if (invert_flag < 0) {
invert_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: invert option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'f':
if (compact_flag < 0) {
compact_flag = 1;
}
else {
say_version();
fprintf(stderr, "error: format-compact option set more than once.\n");
say_help();
return EXIT_FAILURE;
}
break;
case 'h':
say_version();
say_usage();
exit(0);
default:
break;
}
}
if (optind == argc) {
say_version();
fprintf(stderr, "error: not enough arguments.\n");
say_help();
return EXIT_FAILURE;
}
expr = argv[optind++];
if (optind < argc) {
if ((optind == argc - 1) && (input == NULL)) {
input = argv[optind];
}
else {
say_version();
fprintf(stderr, "error: too many options.\n");
say_help();
return EXIT_FAILURE;
}
}
if (count_flag == -1) count_flag = 0;
if (showdist_flag == -1) showdist_flag = 0;
if (showpos_flag == -1) showpos_flag = 0;
if (matchonly_flag == -1) matchonly_flag = 0;
if (showline_flag == -1) showline_flag = 0;
if (invert_flag == -1) invert_flag = 0;
if (compact_flag == -1) compact_flag = 0;
if (dist_flag == -1) dist_flag = 0;
if (verbose_flag == -1) verbose_flag = 0;
if (endline_flag == -1) endline_flag = 0;
if (prefix_flag == -1) prefix_flag = 0;
if (best_flag == -1) best_flag = 0;
if (nondna_flag == -1) nondna_flag = 0;
if (memory_flag == -1) memory_flag = 0;
if (all_flag == -1) all_flag = 0;
if (printline_flag == -1) printline_flag = (!matchonly_flag && !endline_flag && !prefix_flag);
if (!showdist_flag && !showpos_flag && !printline_flag && !matchonly_flag && !showline_flag && !count_flag && !compact_flag && !prefix_flag && !endline_flag) {
say_version();
fprintf(stderr, "Invalid options: No output will be generated.\n");
say_help();
return EXIT_FAILURE;
}
int maskcnt = !count_flag;
int maskinv = !invert_flag * maskcnt;
struct seeqarg_t args;
args.showdist = showdist_flag * maskinv;
args.showpos = showpos_flag * maskinv;
args.showline = showline_flag * maskcnt;
args.printline = printline_flag * maskinv;
args.matchonly = matchonly_flag * maskinv;
args.count = count_flag;
args.compact = compact_flag * maskinv;
args.dist = dist_flag;
args.verbose = verbose_flag;
args.endline = endline_flag * maskinv;
args.prefix = prefix_flag * maskinv;
args.invert = invert_flag * maskcnt;
args.best = best_flag * maskinv;
args.non_dna = nondna_flag;
args.all = all_flag;
args.memory = (size_t)memory_flag * 1024*1024;
return seeq(expr, input, args);
}
int main(int argc, char **argv) {
debug_print("%s (DEBUG)\n", VERSION);
if (argc == 1) {
say_usage();
exit(EXIT_SUCCESS);
}
// Input file names (mock and ChIP).
char *mock_fnames[MAXNARGS+1] = {0};
char *ChIP_fnames[MAXNARGS+1] = {0};
int n_mock_files = 0;
int n_ChIP_files = 0;
int no_mock_specified = 1;
int no_ChIP_specified = 1;
static int list_flag = 0;
static int minmapq = 20;
static int window = 300;
static int mock_flag = 1;
static double minconf = 0.0;
// Needed to check 'strtoul()'.
char *endptr;
// Parse options.
debug_print("%s", "arguments:'\n");
while(1) {
int option_index = 0;
static struct option long_options[] = {
{"chip", required_argument, 0, '1'},
{"confidence", required_argument, 0, 'c'},
{"help", no_argument, 0, 'h'},
{"list-output", no_argument, &list_flag, 1 },
{"mock", required_argument, 0, '0'},
{"no-mock", no_argument, &mock_flag, 0 },
{"quality", required_argument, 0, 'q'},
{"version", no_argument, 0, 'v'},
{"window", required_argument, 0, 'w'},
{0, 0, 0, 0}
};
int c = getopt_long(argc, argv, "0:1:c:hlq:vw:",
long_options, &option_index);
// Done parsing named options. //
if (c == -1) break;
switch (c) {
case 0:
break;
case '0':
debug_print("| mock files(s): %s\n", optarg);
parse_fname(mock_fnames, optarg, &n_mock_files);
no_mock_specified = 0;
break;
case '1':
debug_print("| ChIP files(s): %s\n", optarg);
parse_fname(ChIP_fnames, optarg, &n_ChIP_files);
no_ChIP_specified = 0;
break;
case 'h':
say_usage();
return EXIT_SUCCESS;
case 'l':
list_flag = 1;
break;
case 'c':
// Decode argument with 'strtod()'
errno = 0;
endptr = NULL;
minconf = strtod(optarg, &endptr);
if (!check_strtoX(optarg, endptr) || minconf < 0 || minconf > 1) {
fprintf(stderr,
"zerone error: confidence must be "
"a float between 0 and 1\n");
say_usage();
return EXIT_FAILURE;
}
debug_print("| minconf: %f\n", minconf);
break;
case 'q':
// Decode argument with 'strtoul()'
errno = 0;
endptr = NULL;
minmapq = strtoul(optarg, &endptr, 10);
if (!check_strtoX(optarg, endptr) ||
minmapq < 0 || minmapq > 254) {
fprintf(stderr,
"zerone error: minimum mapping quality must be "
"an integer between 0 and 254\n");
say_usage();
return EXIT_FAILURE;
}
debug_print("| minmapq: %d\n", minmapq);
break;
case 'v':
say_version();
return EXIT_SUCCESS;
case 'w':
window = atoi(optarg);
if (window <= 0) {
fprintf(stderr, "zerone error: window must be a "
"positive integer\n");
say_usage();
return EXIT_FAILURE;
}
debug_print("| window: %d\n", window);
break;
default:
// Cannot parse. //
say_usage();
return EXIT_FAILURE;
}
}
// Now parse positional arguments (file names).
while (optind < argc) {
parse_fname(ChIP_fnames, argv[optind++], &n_ChIP_files);
}
debug_print("%s", "done parsing arguments\n");
// Check options.
if (no_mock_specified && mock_flag) {
fprintf(stderr,
"zerone error: specify a file for mock control experiment\n");
say_usage();
return EXIT_FAILURE;
}
if (no_ChIP_specified) {
fprintf(stderr,
"zerone error: specify a file for ChIP-seq experiment\n");
say_usage();
return EXIT_FAILURE;
}
// Process input files.
zerone_parser_args_t args;
args.window = window;
args.minmapq = minmapq;
ChIP_t *ChIP = parse_input_files(mock_fnames, ChIP_fnames, args);
if (ChIP == NULL) {
fprintf(stderr, "error while reading input\n");
exit(EXIT_FAILURE);
}
// debug info //
{
debug_print("%s", "done reading input files\n");
debug_print("%s", "ChIP:\n");
debug_print("| r = %ld (dimension)\n", ChIP->r);
debug_print("| nb = %d (block number)\n", ChIP->nb);
for (int j = 0 ; j < ChIP->nb ; j++) {
debug_print("| block %s (size: %d)\n",
ChIP->nm + 32*j, ChIP->sz[j]);
}
// Sum reads of all blocks.
size_t *nreads = calloc(ChIP->r, sizeof(size_t));
if (nreads == NULL) {
fprintf(stderr, "memory error\n");
exit(EXIT_FAILURE);
}
for (int i = 0 ; i < nobs(ChIP) ; i++) {
for (int j = 0 ; j < ChIP->r ; j++) {
nreads[j] += ChIP->y[j + i*ChIP->r];
}
}
debug_print("| aggregated mock: %ld reads\n", nreads[0]);
for (int j = 0 ; j < ChIP->r-1 ; j++) {
debug_print("| %s: %ld reads\n", ChIP_fnames[j], nreads[j+1]);
}
free(nreads);
}
// Do zerone.
debug_print("%s", "starting zerone\n");
zerone_t *Z = do_zerone(ChIP);
if (Z == NULL) {
fprintf(stderr, "run time error (sorry)\n");
exit(EXIT_FAILURE);
}
// debug info //
{
debug_print("%s", "Q:\n");
debug_print("%.3f %.3f %.3f\n", Z->Q[0], Z->Q[3], Z->Q[6]);
debug_print("%.3f %.3f %.3f\n", Z->Q[1], Z->Q[4], Z->Q[7]);
debug_print("%.3f %.3f %.3f\n", Z->Q[2], Z->Q[5], Z->Q[8]);
debug_print("%s", "p:\n");
for (int j = 0 ; j < 3 ; j++) {
int off = 0;
char debuf[512];
for (int i = 0 ; i < Z->r+1 ; i++) {
off += sprintf(debuf + off, "%.3f ", Z->p[i+j*(Z->r+1)]);
if (off > 499) break;
}
debug_print("%s\n", debuf);
}
}
// Quality control.
double feat[5];
double QC = zerone_qc(Z, feat);
fprintf(stdout, "# QC score: %.3f\n", QC);
fprintf(stdout, "# features: %.3f, %.3f, %.3f, %.3f, %.3f\n",
feat[0], feat[1], feat[2], feat[3], feat[4]);
fprintf(stdout, "# advice: %s discretization.\n",
QC >= 0 ? "accept" : "reject");
// List output.
if (list_flag) {
int wid = 0;
int target = 0;
double best = 0.0;
for (int i = 0 ; i < ChIP->nb ; i++) {
char *name = ChIP->nm + 32*i;
// Do not print the last bin because it may extend
// beyond the limit of the chromosome.
for (int j = 0 ; j < ChIP->sz[i]-1 ; j++) {
// Toggle on target state.
double conf = Z->phi[2+wid*3];
if (!target && Z->path[wid] == 2 && conf > minconf) {
fprintf(stdout, "%s\t%d\t", name, window*j + 1);
best = conf;
target = 1;
}
// Toggle off target state.
else if (target) {
// Update best score.
if (conf > best) best = conf;
if (Z->path[wid] != 2 || conf < minconf) {
fprintf(stdout, "%d\t%.5f\n", window*(j+1), best);
best = 0.0;
target = 0;
}
}
wid++;
}
// In case the end of the block is a target.
if (target) {
fprintf(stdout, "%d\t%.5f\n", window * ChIP->sz[i], best);
best = 0.0;
target = 0;
}
}
}
// Table output.
else {
// Use 'offset' to navigate in the ChIP blocks.
uint64_t offset = 0;
// In case no mock was provided, skip the column.
const int skipmock = mock_flag ? 0 : 1;
for (int i = 0 ; i < ChIP->nb ; i++) {
char *name = ChIP->nm + 32*i;
// Do not print the last bin because it may extend
// beyond the limit of the chromosome.
for (int j = 0 ; j < ChIP->sz[i]-1 ; j++) {
// Skip if 'confidence' too low.
if (Z->phi[2+(offset+j)*3] < minconf) continue;
fprintf(stdout, "%s\t%d\t%d\t%d", name, window*j + 1,
// Block name, window start, end, state.
window*(j+1), Z->path[offset+j] == 2 ? 1 : 0);
for (int k = skipmock ; k < Z->ChIP->r ; k++) {
fprintf(stdout, "\t%d",
// Read numbers of each file.
Z->ChIP->y[(offset+j)*Z->ChIP->r+k]);
}
fprintf(stdout, "\t%.5f\n",
// Confidence score.
Z->phi[2+(offset+j)*3]);
}
// End of the block. Update 'offset' before
// local window number is reset to 0.
offset += Z->ChIP->sz[i];
}
}
destroy_zerone_all(Z); // Also frees ChIP.
for (int i = 0 ; i < MAXNARGS ; i++) {
free(mock_fnames[i]);
free(ChIP_fnames[i]);
}
return 0;
}
