I have 3 large files with case control data for various SNPs on my gene of interest. The files are in the format .bed, .bim, and .fam. Apparently Haploview only accepts .ped and .info files. Do I need to convert these? What is the best way to move forward with this?
Hello Justine,
you are right. You have to convert the files. This can easily be done in plink and is convenient and fast using PLINK2. Then simply load them into Haploview for visualisation and LD analysis. If your files are really large you may have to set memory allocation for Haploview.
Good luck
Walther.
Usually these 3 files (.bed, .bim, .fam) are the output of plink with the --make-bed option. To see the result with Haploview you first have to continue your association analysis with plink, then visualize the result with Haploview.
In case of a non-stratified population, you do the association analysis with plink --assoc.
In case of a stratified population, you can visualize the clusters (.mds file generated by plink) with haploview, and then perform your association analysis on the stratified population (plink --mh --within .cluster2).
In both cases, you can visualize the output the association analysis with Haploview (better invoke it within gPlink).
Ok thanks. Which files get converted to .ped and which ones get converted to .info? Does each one have to get converted to both? Very sorry if that's a dumb question but I know next to nothing about using bioinformatics programs. I am already having trouble installing PLINK...apparently it requires a working knowledge of command line stuff...
This plink command is going to create the files you need for Haploview
plink --bfile yourfile --recodeHV --out HVfile
In Plink, .bed, .bim and .fam files work together for various data manipulation.
To answer your question, yourfile.bed and yourfile.bim will be converted to HVfile.ped. On the other hand, yourfile.bim is going to be converted to HVfile.info.
I've used --recodeHV in Plink version 1.07. I don't know if this command is working in plink2 (the newest).
bed-bim-fam is Plink's binary (i.e., compressed) format: bed contains genotype data, bim is a map file, and fam is a pedigree file. This format is much more efficient and faster than the standard ped format. The ped file is in linkage format and contains the same pedigree data as the fam file and the genotype information in the bed file. bim and info are both map files.
Plink 1.07 should be fast enough for all but huge datasets (think thousands of individuals imputed to 1,000 genomes). Plink1.9 is amazingly fast, though.
I pasted into Terminal the command "plink --bfile RACI_Case_Control --recodeHV --out HVfile" (my file is RACI_Case_Control.bed)
and it didn't work. I got a response that says "-bash: plink: command not found" What am I doing wrong?
the "--recodeHV" also didn't do anything and gave me the same response..
I had a colleage convert the files to .ped and map and now when I load the file into Haploview I get a fatal error that says
Exception in thread "AWT-EventQueue-0" java.lang.OutOfMemoryError: Java Heap Space
Walter, the file is quite big so I am guessing I need more memory to get Haploview to work? Don't know how to proceed. I have a mac, 2.13 GB RAM. No wonder it failed, the file is larger than my computer's memory...2.9 GB
Java also has memory limitations. You can increase Java's heap space with the -Xmx flag. Then again, if you only have 2 Gb of ram, there's no point.
Is there a reason you want to do analyses in Haploview? It's a great visualization tool, but Plink is much better for association analyses (assuming you're doing run-of-the-mill analyses). If you really want to use Haploview, you can split your files up.
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology