metaharvest inaccurately renames metadata headings retrieved from ENA #27
Description
Currently, metadata fetched from ENA are extracted from the original JSON by metaharvest using the function _extract_ena_metadata, which ruturns a CSV table as output. In this process, headings are renamed, and in particular aliases are remapped as 'geo':
study_alias --> geo_series and sample_alias --> geo_sample. However, this is not always correct, as in those case in which RNA-Seq data have been brockered from Array-Express, or have been submitted directly to (ENA or NCBI) SRA. See, for instance, PRJEB48614
"ena_sample_title","geo_series","geo_sample","ena_project","ena_sample","ena_run","read_count","library_layout","extra"
"1_E_TIGRb_S8","E-MTAB-11129","E-MTAB-11129:1_E_TIGRb_S8","PRJEB48614","SAMEA10792371","ERR7246655","52455546","PAIRED","0"
"1_E_Ua_S1","E-MTAB-11129","E-MTAB-11129:1_E_Ua_S1","PRJEB48614","SAMEA10792373","ERR7246657","52170236","PAIRED","0"
"EPOCE2_T4_2_S21","E-MTAB-11129","E-MTAB-11129:EPOCE2_T4_2_S21","PRJEB48614","SAMEA10792378","ERR7246662","41502100","PAIRED","0"
...
Unfortunately, this change could be breaking...