feat: extract credible sets and studies from eQTL Catalogue finemapping results

[ireneisdoomed]

This PR includes:

• New DAG that generates credible_set and study_index datasets based on the eQTL Catalogue finemapping results for all eQTLs that significantly influences gene expression.
  
• The process takes around 40 minutes and it introduces a novel approach to preprocess input files: instead of handling the compressed TSVs with Spark, we use Dataflow to decompress and store in a temporary bucket so that the ingestion is parallelised.

Metrics

• Nr of studies: 317 911
• Nr of credible sets: 385 100
• Stats about the size of each credible set:

+-------+------------------+                                                    
|summary|       credSetSize|
+-------+------------------+
|  count|            385100|
|   mean|31.648021293170604|
| stddev|  99.3423119691853|
|    min|                 1|
|    25%|                 3|
|    50%|                11|
|    75%|                32|
|    max|              4090|
+-------+------------------+

• Stats about the number of credible sets per study:

+----------------+------+                                                       
|nCredSetPerStudy| count|
+----------------+------+
|              10|     2|
|               9|     3|
|               8|     6|
|               7|    25|
|               6|    71|
|               5|   313|
|               4|  1232|
|               3|  7555|
|               2| 46542|
|               1|262162|
+----------------+------+

Note
This PR only processes results where the quantification method is ge.
Based on a discussion with @DSuveges and @d0choa, we will bring credible sets from all methods. I suggest that these changes happen in a subsequent PR, to checkpoint the work achieved here in case we want to revert.

coauthored with @d0choa

[d0choa]

Wonderful!

[d0choa]

to be added to the study index

[ireneisdoomed]

Yakov mentioned that carrying it over in the study index should be enough.