Some notes on uploading to the European Nucleotide Archive:

This is astoundingly difficult. Here are some notes that may save you the
pain I experienced today (October 2014).

0) Upload your FASTQ, but make sure you upload it as fastq.gz or fastq.bz2.

1) Also upload the MD5 of all those files as blah.fastq.gz.md5

2) only THEN log in to the website, because otherwise it will not know about
your files!

3) First you need to submit a study, which is rather easy

4) Then submit samples. Where it says 'search' for organisms, you can't
enter anything logical like 'MG1655', or an accession ID. Type 'eschericia
coli'.. but slowly. The autocomplete needs to catch up with you. Then try
various variants of the possible name of your organism, and it might come up
with what you need. Or not. 

5) Now you can upload 'runs' to those samples. Note that samples end up in
the sample screen in THE REVERSE ORDER IN WHICH YOU CREATED THEM. If you
gave your samples names, they do NOT appear in the upload form for runs. You
have to copy-paste the accession ID from the sample screen, BUT DO NOTE THE
REVERSE ORDER!

6) If you press 'submit' in the sample upload screen, first nothing happens.
Then a window appears asking you if you are sure you want to submit. This
window subsequently does not go away. If you click 'ok' again, you will get
an error. So after a while click cancel

7) Observe the 'Congratulations!' message and rejoice. You've made it.