Why these steps are necessary
The GATK uses two files to access and safety check access to the reference files: a .dict dictionary of the contig names and sizes and a .fai fasta index file to allow efficient random access to the reference bases. You have to generate these files in order to be able to use a Fasta file as reference.
NOTE: Picard and samtools treat spaces in contig names differently. We recommend that you avoid using spaces in contig names.
Creating the fasta sequence dictionary file
We use CreateSequenceDictionary.jar from Picard to create a .dict file from a fasta file.
1 > java -jar CreateSequenceDictionary.jar R= Homo_sapiens_assembly18.fasta O= Homo_sapiens_assembly18.dict 2 3 [Fri Jun 19 14:09:11 EDT 2009] net.sf.picard.sam.CreateSequenceDictionary R= Homo_sapiens_assembly18.fasta O= Homo_sapiens_assembly18.dict 4 [Fri Jun 19 14:09:58 EDT 2009] net.sf.picard.sam.CreateSequenceDictionary done. 5 Runtime.totalMemory()=2112487424 6 44.922u 2.308s 0:47.09 100.2% 0+0k 0+0io 2pf+0w
Creating the fasta index file
We use the faidx command in samtools to prepare the fasta index file. This file describes byte offsets in the fasta file for each contig, allowing us to compute exactly where a particular reference base at contig:pos is in the fasta file.
1 > samtools faidx Homo_sapiens_assembly18.fasta 2 108.446u 3.384s 2:44.61 67.9% 0+0k 0+0io 0pf+0w
Creating and sorting the vcf index file
1 java -jar picard.jar SortVcf I=test.vcf O=test.sorted.vcf SEQUENCE_DICTIONARY=ucsc.hg19.dict