ANCIENT!

This file contains information about sequence files on the Whitehead CGR computers.

0. Project names.  Most are Ln, where n is a positive integer.  Another is G3.

   Henceforth, let Ln denote a project name, even though not all project names
have this form.

   For each project name, there would be a project directory, but some have been
archived.  Archived projects can be retrieved in about 1 day.

   Individual projects are in /seq/projects (a few) or /seq/projectsn, where n is a
positive integer.

   WARNING: You may have write permission on /seq.  Be careful!

2. Go to /seq/projects.  For n a positive integer, 

   type "cat status_files/Ln.data" to get some information

        look for projecttype = HTGS Regular (as opposed to HTGS Draft)

        look for location, which tells you the organism and chromosome number

   type "gogap Ln" to get to the project directory

        if it's archived, you should discover this now

3. Put yourself in the project directory for Ln, and cd to phrassembly/Phrap.
   Then LnPHRP and LnPHRP.qual contain reads and quality scores, respectively.

   Read labels all(?) have the form >ProjectPlateOther.Instance[.exp] where Project 
   is the name of the project (e.g. L3191), Plate is Pm for some positive integer m, 
   Other is some other stuff, and Instance identifies the attempt to sequence the
   particular DNA fragment (the first is T0, then T1, etc.)  But under another 
   chemistry, one sees 0 instead.  The .exp is there if the read has already been
   trimmed of vector.

   Look at LnPHRP.  Let a,b,c,d denote digits, and let m denote a positive integer.
   >Lines containing Pm where 1 <= m <= 599 are unpaired reads from M13 vectors.
   >Lines containing Pabc[d] where 6 <= a <= 7 are paired reads from plasmid vectors,
    allegedly with a mean insert length of 4000 and a standard deviation of 1000.
   >Lines containing P8bc[d] are paired reads from fosmid vectors,
    allegedly with a mean insert length of 40000 and a standard deviation of 2500.
   Either of the latter two types of lines should have an F or an R in them, for 
   forward or reverse.  Change F to R and look for the changed line.  It is the 
   paired read.

4. Now go to phrassembly/mapper.  Do "grep avg contig.1.txt".  If the project was
   finished, you'll see something like "proj. avg. length: 4753 +/- 652.3".  This
   gives you a priori insert length statistics, which I am assuming represent 
   total lengths of read...read.  

5. All of the above does not get you the "final finishing reads", which are 
   transposon reads.  To get these, go to the project directory for Ln, and look
   for directories which end in _trans.  Cd to one of these, and then cd to Phrap.
   For example, you may now be in L2124/L2124P602H6_trans/Phrap.  In the example,
   the transposon reads are in L2124P602H6PHRP.  However, there is one such file
   for each _tran directory, so in general you'll have a bunch of files of reads.

6. The file "cons" (in the project directory) contains the final contigs.  However, 
   as the project proceeds, these are changed.  I don't know (yet) how to tell if a
   project is COMPLETELY finished.

7. The file "rel" (in the project directory) contains placement information for
   the reads in the final contigs.  It is not clear how seriously it should be
   taken.