Client¶

Preamble¶

In order for the quality control to work optimally, it is recommended to split multiallelic variants. To compare similar variants during the Association Tests, input VCF files need to be annotated with vep (to added the consequence on genes and GnomAD frequencies). The command line to do this (after having installed vep) is:

/path/to/vep --cache --merged --offline --dir [/path/to/cache] --fork 24 --buffer_size 25000 --species homo_sapiens --assembly GRCh37 --use_given_ref --check_existing --allele_number --symbol --af_gnomad --vcf -i input.vcf.gz -o annotated.vcf

Launch the Client GUI¶

java -jar PrivAS.Client.jar [/path/to/data/directory]

Performing Association Tests with PrivAS¶

1. Load Variants¶

Load your annotated VCF file and perform a Quality Control on the Variants and Annotate with GnomAD frequencies [File/Load VCF, Apply QC and Annotate...]. Alternatively if you have already performed this operation for a previous session and you want the same QC parameters, you can simple load the results genotypes file, this will be much quicker [File/Load Annotated QCed Genotypes...]

Choose your input VCF file
Choose a previous QC Parameters file…
Choose a GnomAD File

…or create a new QC Parameters file

The variants are Loaded (as seen in the [Genotype Filename] field)

2. Connect to a RPP Server¶

Connect to a server [Server/Connect to RPP Server...]

Fill in the address
Fill in the port number (default values point to a Server providing the FrEx datasets)

You are connected (as seen in the [RPP Server] field)

3. Perform Association Tests¶

Start a Session : [Server/Start new Session...]

Choose:

one of the provided datasets (your data and RPP’s data must share the same Reference Genome)
the same GnomAD Version as the one you annotated your data with
the least severe VEP Consequence
the maximal frequency in GnomAD
a GnomAD subpopulation (or None to disable)
the maximal frequency in the GnomAD subpopulation
if you want to limit the tests to SNVs (as a INDEL calling is often less reliable)
the Bed file defining the well covered positions (any variants found outside those regions will be ignored)
the QC Parameters file used to extract the Client’s variants, so that the RPP’s variants will be filtered using the same criteria (should be automatically filled in)
the file listing the variants excluded by the QC (should be automatically filled in)
the Association Tests Algorithm (only WSS is present at this time) and its parameters

Set Variants selection criteria and Association Tests parameters

When prompted, save your session (this will allows you to keep track of the parameters you have selected, and to reconnected to the RPP if you have been disconnected).

4. Follow your Association Tests progress¶

Follow the progress of the session in the [Last Known Status] Bar for the RPP server and in the [Third Party Server Log] window

Save your results (when prompted)

5. Visualize your results¶

Save the the visualization (Table / Manhattan Plot) through the [Export] menu

The Main Window¶

Description of the various information found in the Main Window

Field	Description
RPP Server	Reference Panel Provider (RPP) Server The address and port of the RPP Server. The Color indicates the state of the server : Grey : Unknown Green : Up Red : Down
Third Party Server	Third Party Server (TPS) Name of the Server that will perform the actual calculations
Dataset	Name of the reference Dataset
GnomAD Version	Version of the GnomAD Version on the Reference Panel Provider
Max. MAF (GnomAD)	Maximum Minor Allele Frequency Threshold When selecting variants, the Client and the Reference Panel Provider will only keep variants with MAF below or equal to this threshold.
Max. MAF (GnomAD Subpopulation)	Same as above, but for frequencies in Selected Subpopulation
GnomAD Subpopulation	Selected subpopulation on GnomAD
Session ID	Session ID Uniquely identifies your work session for the Client the Reference Panel Provider the Third-Party Server
Least Severe Consequence	Least Severe Consequence When selecting variants, the Client and the Reference Panel Provider will only keep variants with Consequence above or equal to this threshold.
AES Key	AES Key Shared between the Client the Third-Party Server Data exchanged are encrypted/decrypted using this key. Thus the Reference Panel Provider (that serves as a bridge) cannot read these data.
Limit variants to SNVs ?	Limit To SNVs ? When selecting variants, the Client and the Reference Panel Provider will only keep variants that are SNVs.
Algorithm Parameters	Algorithm Parameters The algorithm and parameters that will be used by the Third-Party Server.
Genotype Filename	The Genotype File that was/will be used to extract/hash the data matching selection criteria.
GnomAD filename	GnomAD binary file
Bed of well covered position	Bed file of well covered positions
Hash Key	Hash Key Shared between the Client the Reference Panel Provider This key will be used to hash gene names and variant information, so that the Third-Party Server can do comparison and computing while not being able the read data either from the Client or the Reference Panel Provider.
Public RSA Key	Public RSA Key the Reference Panel Provider uses this key to encrypt the Hash Key, so that it is not legible on the network
Private RSA Key	Private RSA Key the Client use this key to decrypt the Hash Key. Only the Client know this key.
Third Party Public Key	Third Party Public Key This key is used to encrypt your AES key and share it with the Third-Party Server. This encryption prevents the Reference Panel Provider from reading the AES key.
Last Known Status	Last Known Status The Last known message sent by the Reference Panel Provider.
Application Log	Session Log. Messages are in black. Successful operations are in green Errors are in red

PrivAS Client’s Command Lines¶

Main Command¶

Launch the Client GUI¶

java -jar PrivAS.Client.jar [Directory]

[Directory]

initial working directory (Optional, default is current directory)

Tools¶

Perform DEFAULT Quality Control on a VEP Annotated VCF file and convert the result to genotype file¶

java -jar PrivAS.Client.jar vcf2genotypes input.vcf(.gz) GnomADFile.bin

input.vcf(.gz)	The Input VCF file (must have been annotated with vep)
GnomADFile.bin	The GnomAD file to use for the frequency annotation

Perform a Quality Control on a VCF file¶

java -jar PrivAS.Client.jar vcf2qc input.vcf(.gz) qc.param

input.vcf(.gz)	The Input VCF file (must have been annotated with vep)
qc.param	The file containing the QC parameters to apply

Convert a QCed VEP Annotated VCF file to a genotypes file¶

java -jar PrivAS.Client.jar qc2genotypes vep_annotated_QCed_file.vcf(.gz) GnomADFile.bin

vep_annotated_QCed_file.vcf(.gz)	The input VCF File (must result from a PrivAS QC and thus is annotated with vep)
GnomADFile.bin	The GnomAD file to use for the frequency annotation

Creates an annotation binary file from lists of GnomAD (exome/genome) VCF files¶

java -jar PrivAS.Client.jar gnomad gnomADVersion listExomeVCFFiles.list listGenomeVCFFiles.list output.bin

gnomADVersion	The name of the GnomAD Version
listExomeVCFFiles.list	File listing input GnomAD Exome files (one path per line)
listGenomeVCFFiles.list	File listing input GnomAD Genome files (one path per line)
output.bin	The name of the resulting binary file