Questions about UK Biobank "summary_stats" data

[hendrikweisser]

I'm trying to extract phenotype associations for variants in selected genes (~300) from the precomputed UK Biobank GWAS results that you've kindly made available here:
https://github.com/rivas-lab/public-resources/tree/master/uk_biobank/summary_stats
(Thank you for this!)

I have the following questions that I hope you can answer:

1. In the "icdinfo.txt" file, there are different prefixes in the phenotype codes (first column) that seem to refer to different categories (e.g. "BIN", "INI", "HC", ...). Is there any documentation about these prefixes/categories? (I've seen the information in the FAQ here, but it doesn't explain everything.)
2. For some of these phenotype codes, the number at the end identifies the UK Biobank data field for the phenotype, but sometimes "10" or "100" is added in front of the number, and sometimes there's no obvious correspondence to a data field. Again, where can I find more information about the system that was used there?
3. In the same file, what is the meaning of the recurring numbers in columns 2, 4 and 5?
4. Many of the phenotypes listed in "icdinfo.txt" (or visible through the Global Biobank Engine website) are not represented in any of the "icds.sorted.part[n].tsv.gz" files. In particular, phenotypes with codes starting in "BIN", "BIN_FC", "MED" and "QT_FC" are missing. Are the "icds..." files from an older version of the analysis? Could they be updated?

Thanks in advance for your help.