Seqera Platform Feature Requests

Preview is currently supported for the following file types: Nextflow output files ( .command.* , .fusion.* and .exitcode ) Text CSV and TSV PDF HTML Images (JPG, PNG, SVG, etc.) Parquet files are a very common columnar storage format for datasets and preview for this file type should be supported.

There are many cases where you want to use public data but it's not practical or feasible to pull it from the hosted region. In those cases, it would be better to copy the data into your own bucket which will put it alongside your compute environment, saving you costs and time in the long run. An example use case would be copying certain files from the igenomes S3 bucket to your own AWS region or cloud provider for use with nf-core pipelines.

Features such as Data Explorer do not currently allow preview/download/upload for the Launch role and require at least Maintain role permissions. This FR merges all the specific requests.

I am investigating the use of the Seqera Platform for keeping track of processed data sent to AWS S3 via the Nextflow process publishDir directive. I want to query the Seqera Platform/Tower backend to find a given tasks published files and the associated S3 path. I did not see this in the documentation or backend.

To run a pipeline we currently require some variables to be filled in manually with a path to a file(s), which requires basic knowledge of file systems that not everyone has. I have seen the datasets tool where it is possible to upload datasets by clicking through the familiar windows file explorer, but is limited to CSV/TSV files and is not very useful if the file is only going to be used once. Could we combine these two in a more integrative way? The Seqera Platform shows a path to a default file - this way users can still opt to use full-paths. Enable a setting for a button in the nextflow_schema per parameter of choice. Pressing this button opens a file explorer. Selecting a file overrides the full path shown in the Seqera Platform. This idea came from some users who are not familiar with command line and would make running pipelines more convenient for them. Note: We are using an on-premise HPC compute environment, connected via SSH.

Currently, the Nextflow publishDir directive allows users to publish workflow files into a specific folder in cloud storage and (for Amazon S3 only as of writing) associate custom key:value tags with the published files by using the tags option. Adding support in Data Explorer to easily search and filter files in a cloud storage bucket based on these custom tag key:value pairs would allow users to easily locate all files/objects associated with a single workflow, group of workflows, or type of file (eg. plots, analyses, reports).

Add support for previewing HTML files that have embedded images. The Data Explorer preview panel should correctly display any images that have the correctly defined relative link to the image file. Note: this doesn't include other embedded file types (such as Javascript) which will have their own security constraints.