# Gwas analyis on whole exome data without controls **URL:** \ **Category:** Science\ **Created:** [May 25, 2020, 11:30pm UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418 "2020-05-25T23:30:21Z")\ **Posts on this page:** 13\ **Page:** 1
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 25, 2020, 11:30pm UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/1 "2020-05-25T23:30:21Z")
Hello everybody. My name is Vincent and I have a whole exome sequence data for humans(only one phenotype). I’ll like to know SNPs and loci under selection and if they are unique to our population of study.There was not control for the study so I have been advised to compare with exome data from 1000 genomes project. I will be happy if I can be guided on what do. At the moment I have been able to call the snps using bcftools and started practicing some tutorial using hail and got some help from @danking. Thank you in advance. ---
**Author:** ![tpoterba](https://yyz2.discourse-cdn.com/flex036/user_avatar/discuss.hail.is/tpoterba/32/61_2.png) [@tpoterba](https://discuss.hail.is/u/tpoterba)\ **Post date:** [May 26, 2020, 12:37pm UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/2 "2020-05-26T12:37:16Z")
Paging @kumarveerapen - he might be able to help here. I’m guessing that using gnomAD is your best bet here, but I’m not sure exactly how to design the study to avoid confounders. ---
**Author:** ![kumarveerapen](https://yyz2.discourse-cdn.com/flex036/user_avatar/discuss.hail.is/kumarveerapen/32/355_2.png) [@kumarveerapen](https://discuss.hail.is/u/kumarveerapen)\ **Post date:** [May 27, 2020, 1:08am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/3 "2020-05-27T01:08:30Z")
Thanks for tagging me on this, Tim! 🙂 I would first like to clarify about your question, Vincent. You would like to test for SNPs that have undergone selection and are not reported in your reference population? Would you mind us asking: What is the distribution of your population ancestry? I would also follow what Tim has suggested by annotating your variants with our Hail [`annotation_db`](https://hail.is/docs/0.2/annotation_database_ui.html) for gnomAD allele frequencies in order to see what the variant frequencies are. Upon annotation, you could then which variants are relatively rare in your tested population and which are not. ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 27, 2020, 1:57pm UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/4 "2020-05-27T13:57:07Z")
Thanks @tpoterba @kumarveerapen. The population is African Ancestry. Specifically Ghana ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 28, 2020, 4:38am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/5 "2020-05-28T04:38:04Z")
Hi @tpoterba@danking. I attempted annotating my data. First I had to convert to an mt object which was successful thanks to help from @danking. But when I try to annotate with exome\_sites I get the error Traceback (most recent call last): File “”, line 1, in File “”, line 2, in annotate\_rows\_db File “/home/vappiah/.local/lib/python3.6/site-packages/hail/typecheck/check.py”, line 611, in wrapper return _original\_func(\*args, \*\*kwargs_) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/experimental/db.py”, line 209, in annotate\_rows\_db indexed\_value = dataset.index\_compatible\_version(rel.key) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/experimental/db.py”, line 64, in index\_compatible\_version for version in self.versions) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/experimental/db.py”, line 62, in index File “/home/vappiah/.local/lib/python3.6/site-packages/hail/experimental/db.py”, line 64, in for version in self.versions) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/experimental/db.py”, line 29, in maybe\_index return hl.read\_table(self.url).\_maybe\_flexindex\_table\_by\_expr( File “”, line 2, in read\_table File “/home/vappiah/.local/lib/python3.6/site-packages/hail/typecheck/check.py”, line 611, in wrapper return _original\_func(\*args, \*\*kwargs_) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/methods/impex.py”, line 2359, in read\_table for rg\_config in Env.backend().load\_references\_from\_dataset(path): File “/home/vappiah/.local/lib/python3.6/site-packages/hail/backend/spark\_backend.py”, line 327, in load\_references\_from\_dataset return json.loads(Env.hail().variant.ReferenceGenome.fromHailDataset(self.fs.\_jfs, path)) File “/home/vappiah/.local/lib/python3.6/site-packages/py4j/java\_gateway.py”, line 1257, in **call** answer, self.gateway\_client, self.target\_id, self.name) File “/home/vappiah/.local/lib/python3.6/site-packages/hail/backend/spark\_backend.py”, line 40, in deco ‘Error summary: %s’ % (deepest, full, hail. **version** , deepest)) from None hail.utils.java.FatalError: IOException: No FileSystem for scheme: gs Java stack trace: java.io.IOException: No FileSystem for scheme: gs at org.apache.hadoop.fs.FileSystem.getFileSystemClass(FileSystem.java:2660) at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2667) at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:94) at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:2703) at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:2685) at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:373) at org.apache.hadoop.fs.Path.getFileSystem(Path.java:295) at is.hail.io.fs.HadoopFS.fileStatus(HadoopFS.scala:148) at is.hail.io.fs.FS$class.isDir(FS.scala:106) at is.hail.io.fs.HadoopFS.isDir(HadoopFS.scala:57) at is.hail.expr.ir.RelationalSpec$.readMetadata(AbstractMatrixTableSpec.scala:31) at is.hail.expr.ir.RelationalSpec$.readReferences(AbstractMatrixTableSpec.scala:66) at is.hail.variant.ReferenceGenome$.fromHailDataset(ReferenceGenome.scala:587) at is.hail.variant.ReferenceGenome.fromHailDataset(ReferenceGenome.scala) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:498) at py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244) at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357) at py4j.Gateway.invoke(Gateway.java:282) at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132) at py4j.commands.CallCommand.execute(CallCommand.java:79) at py4j.GatewayConnection.run(GatewayConnection.java:238) at java.lang.Thread.run(Thread.java:748) Hail version: 0.2.41-b8144dba46e6 Error summary: IOException: No FileSystem for scheme: gs ---
**Author:** ![danking](https://yyz2.discourse-cdn.com/flex036/user_avatar/discuss.hail.is/danking/32/43_2.png) [@danking](https://discuss.hail.is/u/danking)\ **Post date:** [May 28, 2020, 1:17pm UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/6 "2020-05-28T13:17:56Z")
The annotation DB only works for Hail inside Google Cloud. Hail team maintains [tools to make that easy](https://hail.is/docs/0.2/cloud/google_cloud.html). If you cannot use the cloud, you can download the gnomad loci to your computer from the gnomad website directly: [https://gnomad.broadinstitute.org/downloads](https://gnomad.broadinstitute.org/downloads). Then you can do something like: ```auto gnomad = hl.read_table('/path/to/gnomad.exomes.r2.1.1.sites.ht') mt = mt.annotate_rows(gnomad = gnomad[mt.locus, mt.alleles]) ``` ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 29, 2020, 2:43am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/7 "2020-05-29T02:43:34Z")
Thanks @danking @tpoterba . I was able to download the download the loci and do the annotation. ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 29, 2020, 2:44am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/8 "2020-05-29T02:44:22Z")
@kumarveerapen I was able to do the annotation with my data. Thanks ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 30, 2020, 3:31am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/9 "2020-05-30T03:31:26Z")
@tpoterba @kumarveerapen . The data I have is Ghanaian Ancestry with just 1 phenotype. Because I don’t have controls, I would like to use data from gnomeAD (african subset) as my negative control and then perform the rest of the analysis. I have been able to download the exome data and performed the annotation as you suggested. But i have no idea how to extract only the african population from the data. If you have any other approach i am willing to try as well. Thanks for the help. ---
**Author:** ![kumarveerapen](https://yyz2.discourse-cdn.com/flex036/user_avatar/discuss.hail.is/kumarveerapen/32/355_2.png) [@kumarveerapen](https://discuss.hail.is/u/kumarveerapen)\ **Post date:** [May 30, 2020, 3:45am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/10 "2020-05-30T03:45:04Z")
Thanks for tagging me, Jotes! I would be very weary about using gnomAD as a control dataset because it was curated for the purpose of describing variation and not to be used as a statistical control, especially since the individual phenotype in gnomAD is not known. Moreover, gnomAD individual level data is not publicly available. Are there any other control data that you are able to get your hands on? Say something on dbGAP? That way, you could select a phenotype control closest to the phenotype that you are testing and have improved confidence in your variation identifcation. ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 30, 2020, 4:01am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/11 "2020-05-30T04:01:10Z")
I was planning to use data from 1000 genomes project. But I found out the vcf they have is for whole genome sequence. So I will download the raw exome sequence and call the snps. What do you think of this plan @tpoterba @kumarveerapen @danking ---
**Author:** ![kumarveerapen](https://yyz2.discourse-cdn.com/flex036/user_avatar/discuss.hail.is/kumarveerapen/32/355_2.png) [@kumarveerapen](https://discuss.hail.is/u/kumarveerapen)\ **Post date:** [May 30, 2020, 4:23am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/12 "2020-05-30T04:23:40Z")
That’s a relatively better idea. But I would still implore that you attempt to find a phenotype specific control to improve interpretability of your variants. ---
**Author:** ![jotes](https://avatars.discourse-cdn.com/v4/letter/j/bb73d2/32.png) [@jotes](https://discuss.hail.is/u/jotes)\ **Post date:** [May 30, 2020, 4:35am UTC](https://discuss.hail.is/t/gwas-analyis-on-whole-exome-data-without-controls/1418/13 "2020-05-30T04:35:49Z")
Thanks @kumarveerapen. I will do that.