Hello- Before I reinvent the flat tyre I ask here...
Does anybody know of or have some Java code to make an index file for a fasta file? For "index fasta file" I mean the file produced by e.g. samtools faidx and containing sequence length, byte positions of each sequence etc.
Ideally I'm looking for something working like this:
public static void Main(String[] args){
//Writes to disk index just like `samtools faidx sequences.fasta`:
new MakeFastaIndex("sequences.fasta")
}
I had a look at picard, biojava but couldn't find anything...
2 answers
Without checking anything (lines length...) for text files only:
import java.io.*;
import java.util.*;
public class Faidx
{
public static class SamSequenceRecord
{
String name=null;
int length=0;
long begin;
int linelen=-1;
}
public final List<SamSequenceRecord> dict = new ArrayList<>();
public Faidx(final File fn) throws IOException
{
long offset=0;
FileReader rz = new FileReader(fn);
int c;
SamSequenceRecord ssr = null;
while((c=rz.read())!=-1) {
if (c == '>') {
offset++;
StringBuilder sb=new StringBuilder();
ssr = new SamSequenceRecord();
boolean ws=false;
while((c=rz.read())!=-1 )
{
offset++;
if( c=='\n') break;
if(!ws && Character.isWhitespace(c)) ws=true;
if(!ws) sb.append((char)c);
}
ssr.name=sb.toString();
ssr.begin=offset;
dict.add(ssr);
}
else
{
int n=0;
do {
offset++;
if( c == '\n') break;
n++;
c = rz.read();
} while(c!=-1);
ssr.length+=n;
if(ssr.linelen==-1)
{
ssr.linelen = n;
}
}
}
}
public void print(PrintStream out)
{
for(SamSequenceRecord ssr: dict)
{
out.printlnssr.name+"\t"+ssr.length+"\t"+ssr.begin+"\t"+ssr.linelen+"\t"+(ssr.linelen+1));
}
}
public static void main(String args[]) throws IOException {
new Faidx(new File(args[0])).print(System.out);
}
}
If anyone is interested I put here https://github.com/dariober/ASCIIGenome/tree/master/src/faidx a package to index a fasta file. Some tests are https://github.com/dariober/ASCIIGenome/blob/master/test/faidx.
Example:
new Faidx(new File("seq.fa"));
will create seq.fa.fai
Great project Dario (ASCIIGenome)!
Log in to answer this question.
-
Translating DNA to protein using biojava
written by delcophantom8491 •
-
Extracting sequence, not reads, by position from bam
written by geneware •
-
Corrupted tabix index?
written by dariober
-
Reading FASTA with BioJava 4.0.0
written by priess1991 •
-
Fetching Description And Accession Number From A Genbank Format Dna Sequence File Using Biojava
written by J.Ashley •
-
Parsing A Fasta File In Java
written by J.Ashley •
-
Generating Sequence From Pdb File Using Biojava
written by potassiumiodide0990
-
To Parse A Fasta File And Retrieve Only The Raw Sequence Of A Specific Chain
written by priyankapillai13 •
-
Extracting Chain Specific Fasta Sequence From A Fasta.Txt File Using Java/Biojava
written by potassiumiodide0990
-
Help Me To Read The Fasta File Through Biojava.
written by Muhammad •
I would assume that there is something in htsjdk for this.
@devon I didn't find it.
What about "dict"?
'Dict' is for sequence dictionary. The structure of a dict is different from a faidx (it contains a SAM sequence header dictionary)
Ah, right. I'm a little surprised that this isn't in htsjdk, seems like a normal function for it to provide.