The genetic map

It is strongly recommended to provide to ShapeIT a genetic map of the chromosome you want to phase. It will greatly improve accuracy of phasing. You can find genetic maps for Humans here:

• HapMap phase II NCBI b36
• HapMap phase II NCBI b37

The genetic map provided to ShapeIT should have 3 columns delimited by a single space character. The first column is the physical position (in bp), the second one is the rate (in cM/Mb) and the third one is the genetic position (in cM). Only the first and the third columns are used by ShapeIT, so the second on can be set to any values if you generate this file on your own. The HapMap phase II NCBI b36 genetic map is already formatted correctly for ShapeIT. If you use the HapMap phase II NCBI b37, you have to change the format of the file in order to use it with ShapeIT. To change the format, use the following command:

for chr in $seq(1 22); do cat genetic_mapGRC_chr$chr.txt | awk '{ print $2, $3, $4}' > genetic_mapGRC_chr$chr_reformatted.txt; done Note: The physical positions in the snp map file (MAP or BIM or INP) and in the genetic map file MUST be of the same NCBI version. If your positions in the BIM/MAP file of your data are from the b36, you MUST use the HapMap phase II b36 genetic map! Note: Do not provide the genetic positions as part of the MAP/BIM file as it is possible to do with Plink. ShapeIT do no read this field in those files.