|
| 1 | +#include "DBReader.h" |
| 2 | +#include "DBWriter.h" |
| 3 | +#include "Debug.h" |
| 4 | +#include "Util.h" |
| 5 | +#include "MemoryMapped.h" |
| 6 | +#include "Orf.h" |
| 7 | +#include "Parameters.h" |
| 8 | + |
| 9 | +#ifdef OPENMP |
| 10 | +#include <omp.h> |
| 11 | +#endif |
| 12 | + |
| 13 | +int gff2db(int argc, const char **argv, const Command &command) { |
| 14 | + Parameters &par = Parameters::getInstance(); |
| 15 | + par.parseParameters(argc, argv, command, true, Parameters::PARSE_VARIADIC, 0); |
| 16 | + |
| 17 | + std::string outDb = par.filenames.back(); |
| 18 | + par.filenames.pop_back(); |
| 19 | + std::string seqDb = par.filenames.back(); |
| 20 | + par.filenames.pop_back(); |
| 21 | + |
| 22 | + DBReader<unsigned int> reader(seqDb.c_str(), (seqDb + ".index").c_str(), par.threads, DBReader<unsigned int>::USE_INDEX | DBReader<unsigned int>::USE_DATA | DBReader<unsigned int>::USE_LOOKUP_REV); |
| 23 | + reader.open(DBReader<unsigned int>::NOSORT); |
| 24 | + DBReader<unsigned int> headerReader((seqDb + "_h").c_str(), (seqDb + "_h.index").c_str(), par.threads, DBReader<unsigned int>::USE_INDEX | DBReader<unsigned int>::USE_DATA); |
| 25 | + headerReader.open(DBReader<unsigned int>::NOSORT); |
| 26 | + |
| 27 | + std::string outDbIndex = outDb + ".index"; |
| 28 | + DBWriter writer(outDb.c_str(), outDbIndex.c_str(), par.threads, par.compressed, Parameters::DBTYPE_NUCLEOTIDES); |
| 29 | + writer.open(); |
| 30 | + std::string outHdr = outDb + "_h"; |
| 31 | + std::string outHdrIndex = outDb + "_h.index"; |
| 32 | + DBWriter headerWriter(outHdr.c_str(), outHdrIndex.c_str(), par.threads, par.compressed, Parameters::DBTYPE_GENERIC_DB); |
| 33 | + headerWriter.open(); |
| 34 | + std::string outLookup = outDb + ".lookup"; |
| 35 | + std::string outLookupIndex = outDb + ".lookup.index"; |
| 36 | + DBWriter lookupWriter(outLookup.c_str(), outLookupIndex.c_str(), par.threads, par.compressed, Parameters::DBTYPE_OMIT_FILE); |
| 37 | + lookupWriter.open(); |
| 38 | + |
| 39 | + FILE *source = FileUtil::openAndDelete((outDb + ".source").c_str(), "w"); |
| 40 | + for (size_t i = 0; i < par.filenames.size(); ++i) { |
| 41 | + if (fprintf(source, "%zu\t%s\n", i, FileUtil::baseName(par.filenames[i]).c_str()) < 0) { |
| 42 | + Debug(Debug::ERROR) << "Cannot write to file " << outDb << ".source\n"; |
| 43 | + return EXIT_FAILURE; |
| 44 | + } |
| 45 | + } |
| 46 | + if (fclose(source) != 0) { |
| 47 | + Debug(Debug::ERROR) << "Cannot close file " << outDb << ".source\n"; |
| 48 | + return EXIT_FAILURE; |
| 49 | + } |
| 50 | + |
| 51 | + const std::vector<std::string> features = Util::split(par.gffType, ","); |
| 52 | + if (features.empty()) { |
| 53 | + Debug(Debug::WARNING) << "No feature types given. All features will be extracted\n"; |
| 54 | + } |
| 55 | + std::vector<size_t> featureCount(features.size(), 0); |
| 56 | + |
| 57 | + if (par.filenames.size() < reader.getSize()) { |
| 58 | + Debug(Debug::WARNING) << "Not enough GFF files are provided. Some results might be omitted\n"; |
| 59 | + } |
| 60 | + |
| 61 | + unsigned int entries_num = 0; |
| 62 | + Debug::Progress progress(par.filenames.size()); |
| 63 | +#pragma omp parallel |
| 64 | + { |
| 65 | + int thread_idx = 0; |
| 66 | +#ifdef OPENMP |
| 67 | + thread_idx = omp_get_thread_num(); |
| 68 | +#endif |
| 69 | + |
| 70 | + char buffer[32768]; |
| 71 | + const char* fields[255]; |
| 72 | + |
| 73 | + std::string revStr; |
| 74 | + revStr.reserve(par.maxSeqLen + 1); |
| 75 | + |
| 76 | + std::vector<size_t> localFeatureCount(features.size(), 0); |
| 77 | + |
| 78 | +#pragma omp for schedule(dynamic, 1) nowait |
| 79 | + for (size_t i = 0; i < par.filenames.size(); ++i) { |
| 80 | + progress.updateProgress(); |
| 81 | + MemoryMapped file(par.filenames[i], MemoryMapped::WholeFile, MemoryMapped::SequentialScan); |
| 82 | + if (!file.isValid()) { |
| 83 | + Debug(Debug::ERROR) << "Could not open GFF file " << par.filenames[i] << "\n"; |
| 84 | + EXIT(EXIT_FAILURE); |
| 85 | + } |
| 86 | + char *data = (char *) file.getData(); |
| 87 | + char* end = data + file.mappedSize(); |
| 88 | + size_t idx = 0; |
| 89 | + while (data < end && *data != '\0') { |
| 90 | + // line is a comment or empty |
| 91 | + if (*data == '#' || *data == '\n') { |
| 92 | + data = Util::skipLine(data); |
| 93 | + continue; |
| 94 | + } |
| 95 | + |
| 96 | + const size_t columns = Util::getFieldsOfLine(data, fields, 255); |
| 97 | + data = Util::skipLine(data); |
| 98 | + if (columns < 9) { |
| 99 | + Debug(Debug::WARNING) << "Not enough columns in GFF file\n"; |
| 100 | + continue; |
| 101 | + } |
| 102 | + |
| 103 | + if (features.empty() == false) { |
| 104 | + bool shouldSkip = true; |
| 105 | + std::string type(fields[2], fields[3] - fields[2] - 1); |
| 106 | + for (size_t i = 0; i < features.size(); ++i) { |
| 107 | + if (type.compare(features[i]) == 0) { |
| 108 | + localFeatureCount[i]++; |
| 109 | + shouldSkip = false; |
| 110 | + break; |
| 111 | + } |
| 112 | + } |
| 113 | + if (shouldSkip) { |
| 114 | + continue; |
| 115 | + } |
| 116 | + } |
| 117 | + size_t start = Util::fast_atoi<size_t>(fields[3]); |
| 118 | + size_t end = Util::fast_atoi<size_t>(fields[4]); |
| 119 | + if (start == end) { |
| 120 | + Debug(Debug::WARNING) << "Invalid sequence length in line " << idx << "\n"; |
| 121 | + continue; |
| 122 | + } |
| 123 | + std::string strand(fields[6], fields[7] - fields[6] - 1); |
| 124 | + std::string name(fields[0], fields[1] - fields[0] - 1); |
| 125 | + size_t lookupId = reader.getLookupIdByAccession(name); |
| 126 | + if (lookupId == SIZE_MAX) { |
| 127 | + Debug(Debug::ERROR) << "GFF entry not found in database lookup: " << name << "\n"; |
| 128 | + EXIT(EXIT_FAILURE); |
| 129 | + } |
| 130 | + unsigned int lookupKey = reader.getLookupKey(lookupId); |
| 131 | + size_t seqId = reader.getId(lookupKey); |
| 132 | + if (seqId == UINT_MAX) { |
| 133 | + Debug(Debug::ERROR) << "GFF entry not found in sequence database: " << name << "\n"; |
| 134 | + EXIT(EXIT_FAILURE); |
| 135 | + } |
| 136 | + |
| 137 | + unsigned int key = __sync_fetch_and_add(&(entries_num), 1); |
| 138 | + size_t bufferLen; |
| 139 | + if (strand == "+") { |
| 140 | + bufferLen = Orf::writeOrfHeader(buffer, lookupKey, start, end, 0, 0); |
| 141 | + } else { |
| 142 | + bufferLen = Orf::writeOrfHeader(buffer, lookupKey, end, start, 0, 0); |
| 143 | + } |
| 144 | + headerWriter.writeData(buffer, bufferLen, key, thread_idx); |
| 145 | + |
| 146 | + char *seq = reader.getData(seqId, thread_idx); |
| 147 | + //check the strand instead of end - start |
| 148 | + size_t length = end - start + 1; |
| 149 | + writer.writeStart(thread_idx); |
| 150 | + if (strand == "+") { |
| 151 | + size_t len = snprintf(buffer, sizeof(buffer), "%u\t%s_%zu_%zu_%zu\t%zu\n", key, name.c_str(), idx, start, end, i); |
| 152 | + lookupWriter.writeData(buffer, len, thread_idx, false, false); |
| 153 | + writer.writeAdd(seq + start - 1 , length, thread_idx); |
| 154 | + } else { |
| 155 | + size_t len = snprintf(buffer, sizeof(buffer), "%u\t%s_%zu_%zu_%zu\t%zu\n", key, name.c_str(), idx, end, start, i); |
| 156 | + lookupWriter.writeData(buffer, len, thread_idx, false, false); |
| 157 | + // safe accessing in reverse order |
| 158 | + for (size_t j = 0; j < length; j++) { |
| 159 | + revStr.append(1, Orf::complement(seq[end - 1 - j])); |
| 160 | + } |
| 161 | + |
| 162 | + writer.writeAdd(revStr.c_str(), revStr.size(), thread_idx); |
| 163 | + revStr.clear(); |
| 164 | + } |
| 165 | + writer.writeAdd("\n", 1, thread_idx); |
| 166 | + writer.writeEnd(key, thread_idx); |
| 167 | + idx++; |
| 168 | + } |
| 169 | + file.close(); |
| 170 | + } |
| 171 | +#pragma omp critical |
| 172 | + for (size_t i = 0; i < features.size(); ++i) { |
| 173 | + featureCount[i] += localFeatureCount[i]; |
| 174 | + } |
| 175 | + } |
| 176 | + lookupWriter.close(true); |
| 177 | + FileUtil::remove(lookupWriter.getIndexFileName()); |
| 178 | + headerWriter.close(true); |
| 179 | + writer.close(true); |
| 180 | + headerReader.close(); |
| 181 | + reader.close(); |
| 182 | + if (Debug::debugLevel >= Debug::INFO && features.size() > 0) { |
| 183 | + Debug(Debug::INFO) << "Found these feature types and counts:\n"; |
| 184 | + for (size_t i = 0; i < features.size(); ++i) { |
| 185 | + Debug(Debug::INFO) << " - " << features[i] << ": " << featureCount[i] << "\n"; |
| 186 | + } |
| 187 | + } else { |
| 188 | + Debug(Debug::INFO) << (entries_num + 1) << " features were extracted\n"; |
| 189 | + } |
| 190 | + |
| 191 | + if (par.filenames.size() > 1 && par.threads > 1) { |
| 192 | + // make identifiers stable |
| 193 | +#pragma omp parallel |
| 194 | + { |
| 195 | +#pragma omp single |
| 196 | + { |
| 197 | +#pragma omp task |
| 198 | + { |
| 199 | + DBWriter::createRenumberedDB(outHdr, outHdrIndex, "", ""); |
| 200 | + } |
| 201 | + |
| 202 | +#pragma omp task |
| 203 | + { |
| 204 | + DBWriter::createRenumberedDB(outDb, outDbIndex, outDb, outDbIndex); |
| 205 | + } |
| 206 | + } |
| 207 | + } |
| 208 | + } |
| 209 | + |
| 210 | + return EXIT_SUCCESS; |
| 211 | +} |
0 commit comments