Hi all,
We store our alignment files on aws s3. I would like to be able to open them with IGV without needing to download them completely, but I can't find an optimal solution.
If I get a pre-signed url it works but it's not convenient.
I try to follow this but I can't do it at all. I think the problem comes from the configuration file at the end, I don't know exactly what is needed or not. Knowing that we do not have a connection by google, I would just like to use our aws identifiers. When i click on login on igv and add my longin and password i receive a "succes" answer; but then nothing happen on igv, the menu "amazon --> Load from s3 bucket is still inaccessible.
i use this json (i replace the XXXX by the corresponding value):
{
"apiKey": "",
"project_id": "igv",
"auth_provider": "Amazon",
"aws_region": "us-east-1",
"scope": "email",
"redirect_uris": [
"http://localhost:60151/oauthCallback"
],
"client_id": "XXXXXX",
"client_secret": "XXXXX",
"authorization_endpoint": "XXXXXX",
"token_endpoint": "XXXX",
"aws_cognito_pool_id": "XXXX",
"aws_cognito_fed_pool_id": "",
"aws_cognito_role_arn": "XXXXX"
}
It's been on my roadmap to upgrade our current solution to using federated access. For now, we have all the alignments stored in a public S3 bucket but restrict access to certain IP addresses and VPC endpoints. It works for both IGV (end-users need to be connected via VPN) and our internal genome browser.
To limit access on S3 buckets: https://aws.amazon.com/premiumsupport/knowledge-center/block-s3-traffic-vpc-ip/
May want to ask this on IGV support forum: https://groups.google.com/g/igv-help