removed historically interesting, but not terribly relevant summer 1999 minutes

minutes/02071999

-2nd July 1999
--------------
-
-Tony, Nicole, Henning, Tim, Michele
-Peter.
-
-
-Nicole has started putting fasta on ftp site. 
-
-
-Need increments so we can we synchronize. - but we don't know ehre the
-increments start from.
-
-Nicole is going to dump everything this weekend then start adding increments so we know where we are.  Tim wants all human genomic and Greg Schuler's list. 
-
-Nicole can't generate an sql query that regenerates Greg's list.
-
-Tim wants to have a list of things that are in our human unfinished but not (yet) in Greg's list.
-
-
-Web site - T. says we should be using mod-perl - talk next week.
-
-
-Henning wants embl files to tremblize - he needs to talk to us (Ewan,James) about what he needs in the embl files.  He's only going to process CDS not exons.
-
-
-Representing frameshifts in embl files. Can't represent insertions.  Someone (?) is working on this (by 1st sept by agreement with ncbi) but she's not working on this.   There is some  translational exception code in embl where you can say bases 5-8 code for such and such an amino acid.  We could use this to cover frameshifts.
-
-Henning's trembl instructions. 
-------------------------------
-
-These are on the web.
-
-Currently :
-
-EMBL -> trembl.c -> flatfile tremblnew -> automatic annotation.
-
-future -  embl to swissprot
-
-Tags essential
-
-protein  id - id and acc.
-DR for orginal embl acc -(with version no.)
-
-Description line - product, standard name, gene, note - lots of garbage here.
-                   E.C number. Also the DE line from the EMBL entry.
-
-most important are product, E.C. no.
-
-
-If no product information but gene information do something like: 
-
-   Hypothetical 41.9k protein - genename (fragment).
-
-OS line - OC lines generated from this.
-
-
-web sites are :
-
-TrEMBLnew doc is placed in:
-http://www3.ebi.ac.uk/~sp/intern/projects/trembl/guide/index.html
-
-TrEMBLnew entry example:
-http://www3.ebi.ac.uk/~martin/TrEMBLnew.html
-
--- 

minutes/09071999

-9th July 1999
--------------
-
-Peter
-Ewan
-James
-Guy
-Tony
-Miriam
-
-
-Nothing new to report.  Tony talked about mod-perl
-
-We site www.apache.org for details of porting
-

minutes/16081999

-Ewan
-Michele
-James
-Peter
-Elia
-Alessandro
-
-Michele reviewed Ewan's database access code.
-
-Obj.pm - locks on write code
-
-Checking for existing genes before writing.
-
-On updated genes we decided to keep everything and add two tags into the database.  Version and a status (current/obsolete) tag for Gene, Transcript and Exon.
-
-Need write_Contig and write_Clone methods (Michele)
-
-Sequence caching - could end up with the whole database in a cache.  We probably need to have a database connected Object for sequence like Clone and Contig.
-
-Translating exons - The transcript object assumes that the sequence is continuous and all exon phases are inferred from the first exon.  This is not going to be true for non continuous sequence.
-
-
-

minutes/24061999

-Nicole
-Tim
-Me
-Peter S.
-Ewan
-Pat.
-
-
-Original dna sequence - in clone directories.
-
-EMBL file generation - Ewan is going to use BioPerl.
-T. Should have same code for Embl generation. James has
-   code for fast EMBL file generation.
-
-N.  Can dump out fasta - some problems in post-processing.
-T.  How about unfinished to finished - we have to deal with this.
-    As the AC stays the same - can merge finished and unfinished.
-
-
-E. Exception hndling.
-
-
-rethrowing - grep on the $@
-if it isn't the error you want 
-then die($@);
-
-exceptions use the variable name - can put info in 
-here on how the object was generated.  Will be seen
-in exception output.
-
-
-Pat. Asking why she was asked to look at locuslink.
-   
-E.   We need a map. Qustions need to be answered like - which clones
-     are between two markers (and which genes).
-
-Pat. Can already ask for STSs between two markers.  
-
-N.   How to we get genome sequence onto map?
-
-T.   Use ePCR?
-
-E.   What is the datastructure?
-
-Pat. RDB tables. E. want to have a look at these.
-
-E.   Emphasize we need genetic map data as it most useful.
-
-T.   Emphasizes that maybe we should invert the accepted way of doing thins
-     way the map is seen as 'correct' and sequence is fitted to it and use the 
-    map info as evidence when piecing together sequences.
-
-
-
-
-
-