DuckDB compatibility with HPC resources

[apigott]

Not all HPCs (ex. ZIB) are set up to work with DuckDB download settings. Proposed fix TBD

Current temp fixes are:

1. Manual download and snakemake --touch to force use of manually downloaded files
2. Adding proxy settings. Specifically for ZIB cluster:

duckdb.query("SET http_proxy='squid.zib.de:3128';")```

http_proxy could be parameterized from the environment variable `http_proxy` and this is probably the best long term solution to be implemented in a PR

[trevorb1]

I've had issues with specific Linux versions causing duckdb issues before. See #611. Not sure if its related