X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=Transcripts.h;h=2637d35d197b0070e636a16a0404c669190b6e6d;hb=97554bbac838f2ed578d81f98e421dac0669e74e;hp=2750e0c36394d3c3be5cb7d63dfb74f7ae5687a3;hpb=a95154919f950f86de9104b2b9dcf1f0c7e83387;p=rsem.git

diff --git a/Transcripts.h b/Transcripts.h
index 2750e0c..2637d35 100644
--- a/Transcripts.h
+++ b/Transcripts.h
@@ -10,7 +10,10 @@
 #include<fstream>
 #include<vector>
 #include<algorithm>
+#include<map>
+#include<string>
 
+#include "my_assert.h"
 #include "Transcript.h"
 
 
@@ -20,6 +23,8 @@ public:
 		M = 0; this->type = type;
 		transcripts.clear();
 		transcripts.push_back(Transcript());
+
+		e2i.clear(); i2e.clear();
 	}
 
 	int getM() { return M; }
@@ -42,9 +47,23 @@ public:
 	void readFrom(const char*);
 	void writeTo(const char*);
 
+	//Eid: external sid
+	int getInternalSid(int eid) {
+		assert(eid > 0 && eid <= M);
+		return e2i[eid];
+	}
+
+	const Transcript& getTranscriptViaEid(int eid) {
+		return transcripts[getInternalSid(eid)];
+	}
+
+	void buildMappings(int, char**);
+
 private:
 	int M, type; // type 0 from genome , 1 standalone transcriptome
 	std::vector<Transcript> transcripts;
+
+	std::vector<int> e2i, i2e; // external sid to internal sid, internal sid to external sid
 };
 
 void Transcripts::readFrom(const char* inpF) {
@@ -55,8 +74,7 @@ void Transcripts::readFrom(const char* inpF) {
 
 	fin>>M>>type;
 	getline(fin, line);
-	transcripts.clear();
-	transcripts.resize(M + 1);
+	transcripts.assign(M + 1, Transcript());
 	for (int i = 1; i <= M; i++) {
 		transcripts[i].read(fin);
 	}
@@ -72,4 +90,30 @@ void Transcripts::writeTo(const char* outF) {
 	fout.close();
 }
 
+void Transcripts::buildMappings(int n_targets, char** target_name) {
+	std::map<std::string, int> dict;
+	std::map<std::string, int>::iterator iter;
+
+	general_assert(n_targets == M, "Number of transcripts does not match! Please check if the reads are aligned to a transcript set (instead of a genome)!");
+
+	dict.clear();
+	for (int i = 1; i <= M; i++) {
+		const std::string& tid = transcripts[i].getTranscriptID();
+		iter = dict.find(tid);
+		assert(iter == dict.end());
+		dict[tid] = i;
+	}
+
+	e2i.assign(M + 1, 0);
+	i2e.assign(M + 1, 0);
+	for (int i = 0; i < n_targets; i++) {
+		iter = dict.find(std::string(target_name[i]));
+		general_assert(iter != dict.end(), "RSEM can not recognize transcript " + cstrtos(target_name[i]) + "!");
+		general_assert(iter->second > 0, "Reference sequence name " + cstrtos(target_name[i]) + " is duplicated!");
+		e2i[i + 1] = iter->second;
+		i2e[iter->second] = i + 1;
+		iter->second = -1;
+	}
+}
+
 #endif /* TRANSCRIPTS_H_ */