void makeConfigFromTabList(char *input, char *output, boolean useTarget)
/* makeConfigFromFileList - Create config file for hgBedsToBedExps from list of file/cell/ab
or file/cell/ab/target. */
{
struct lineFile *lf = lineFileOpen(input, TRUE);
char *row[4];
FILE *f = mustOpen(output, "w");
while (lineFileRow(lf, row))
{
char *fileName = row[0];
char *cell = row[1];
char *factor = row[2];
verbose(3, "%s\n", fileName);
if (useTarget)
// 4 column input file -- output target cell+treatment+factor
fprintf(f, "%s\t%s+%s\t", row[3], cell, factor);
else
// antibody cell+treatment
fprintf(f, "%s\t%s", factor, cell);
fprintf(f, "\t%s\t", cellAbbreviation(cell));
fprintf(f, "file\t%d\t", scoreCol-1);
fprintf(f, "%g\t", calcNormScoreFactor(fileName, scoreCol-1));
fprintf(f, "%s\n", fileName);
}
lineFileClose(&lf);
carefulClose(&f);
}
void regClusterMakeTableOfTables(char *type, char *input, char *output)
/* regClusterMakeTableOfTables - Make up a table of tables for regCluster program. */
{
FILE *f = mustOpen(output, "w");
struct slName *in, *inList = readAllLines(input);
/* Generally we'll have a bunch of file names that all start and/or end with the same
* thing. This loop will isolate out the bits that vary, and then call a type-specific
* routine to output the metadata columns from the middle parts. */
int commonPrefix = commonPrefixSize(inList);
int commonSuffix = commonSuffixSize(inList);
uglyf("regClusterMakeTableOfTables(type=%s, input=%s, output=%s)\n", type, input, output);
int scoreIx = scoreColIx - 1;
for (in = inList; in != NULL; in = in->next)
{
verbose(2, "Processing %s\n", in->name);
fprintf(f, "%s\t0\t1\t2\t%d\t", in->name, scoreIx);
fprintf(f, "%g", calcNormScoreFactor(in->name, scoreIx));
char *s = in->name;
int len = strlen(s);
char *midString = cloneStringZ(s+commonPrefix, len - commonPrefix - commonSuffix);
if (sameString(type, "uw01"))
uw01MetaOut(f, midString);
else if (sameString(type, "uw02"))
uw02MetaOut(f, midString);
else if (sameString(type, "ans01"))
ans01MetaOut(f, midString);
else if (sameString(type, "ans02"))
ans02MetaOut(f, midString);
else if (sameString(type, "enh01"))
enh01MetaOut(f, midString);
else if (sameString(type, "awgDnase01"))
awgDnase01MetaOut(f, midString);
else
errAbort("Unknown type '%s' in first command line parameter.", type);
freez(&midString);
fprintf(f, "\n");
}
carefulClose(&f);
}
void makeConfigFromFileList(char *input, char *output)
/* makeConfigFromFileList - Create config file for hgBedsToBedExps from list of files.. */
{
FILE *f = mustOpen(output, "w");
struct slName *in, *inList = readAllLines(input);
int commonPrefix = commonPrefixSize(inList);
int commonSuffix = commonSuffixSize(inList);
for (in = inList; in != NULL; in = in->next)
{
char *s = in->name;
int len = strlen(s);
char *midString = cloneStringZ(s+commonPrefix, len - commonPrefix - commonSuffix);
char *factor, *cell;
camelParseTwo(midString, &cell, &factor);
fprintf(f, "%s\t%s\t", factor, cell);
fprintf(f, "%s\t", cellAbbreviation(cell));
fprintf(f, "file\t%d\t", scoreCol-1);
fprintf(f, "%g\t", calcNormScoreFactor(in->name, scoreCol-1));
fprintf(f, "%s\n", in->name);
}
carefulClose(&f);
}
void makeConfigFromEncodeList(char *input, char *output)
/* create config file for hgBedsToBedExps from tab-separated file of format
* <relDate> <fileName> <fileSize> <submitDate> <metadata> */
{
FILE *f = mustOpen(output, "w");
struct lineFile *lf = lineFileOpen(input, TRUE);
char *line;
while (lineFileNextReal(lf, &line))
{
/* Parse out line into major components. */
char *releaseDate = nextWord(&line);
char *fileName = nextWord(&line);
char *fileSize = nextWord(&line);
char *submitDate = nextWord(&line);
char *metadata = trimSpaces(line);
if (isEmpty(metadata))
errAbort("line %d of %s is truncated", lf->lineIx, lf->fileName);
verbose(2, "releaseDate=%s; fileName=%s; fileSize=%s; submitDate=%s; %s\n",
releaseDate, fileName, fileSize, submitDate, metadata);
/* Loop through metadata looking for cell and antibody. Metadata
* is in format this=that; that=two words; that=whatever */
char *cell = NULL, *antibody = NULL;
for (;;)
{
/* Find terminating semicolon if any replace it with zero, and
* note position for next time around loop. */
metadata = skipLeadingSpaces(metadata);
if (isEmpty(metadata))
break;
char *semi = strchr(metadata, ';');
if (semi != NULL)
*semi++ = 0;
/* Parse out name/value pair. */
char *name = metadata;
char *value = strchr(metadata, '=');
if (value == NULL)
errAbort("Missing '=' in metadata after tag %s in line %d of %s",
name, lf->lineIx, lf->fileName);
*value++ = 0;
name = trimSpaces(name);
value = trimSpaces(value);
/* Look for our tags. */
if (sameString(name, "cell"))
cell = value;
else if (sameString(name, "antibody"))
antibody = value;
metadata = semi;
}
if (cell == NULL)
errAbort("No cell in metadata line %d of %s", lf->lineIx, lf->fileName);
if (antibody == NULL)
errAbort("No antibody in metadata line %d of %s", lf->lineIx, lf->fileName);
fprintf(f, "%s\t%s\t", antibody, cell);
fprintf(f, "%s\t", cellAbbreviation(cell));
fprintf(f, "file\t%d\t", scoreCol-1);
fprintf(f, "%g", calcNormScoreFactor(fileName, scoreCol-1));
fprintf(f, "\t%s\n", fileName);
}
carefulClose(&f);
}