!dros_sequence_set.README.v3.87
!October 25 2003


        TRAINING AND SCREENING SETS FOR DROSOPHILA MELANOGASTER.
        ========================================================

These sets of sequences were compiled by Takis Benos (EBI), Leyla Bayraktaroglu
(Harvard) and Michael Ashburner (EBI & Cambridge) with help from Aubrey de Grey
(Cambridge), Joe Chillemi (Harvard) and Martin Reese (LBL). We thank Suzi Lewis
(Berkeley) for inspiration and discussion, and Guochun Liao (Berkeley) for his
repeat sequence set and newly discovered transposable element sequences
from the Berkeley P1 clones.

These sequence sets have been designed to provide either training sets for gene
prediction or screening sets for genomic sequences for Drosophila melanogaster
genomic sequence data.

These files are available from:

ftp://ftp.ebi.ac.uk/pub/databases/edgp/sequence_sets/

nuclear_cds_set
mitochondrial_cds_set
repeat_sequence_set
transposon_sequence_set

unique-50-genes.80.fa
unique-50-genes.50.fa

files with .embl extension are in EMBL format.
files with .fa extension are in FASTA format.

[nuclear_cds.embl.v1.5 was the last released made only by Method 1,
see below. It will remain available.  After v1.5 all nuclear_cds.embl
sets are combined sets made by Method 1 and Method 2.]

The directory:
ftp://ftp.ebi.ac.uk/pub/databases/edgp/sequence_sets/ARCHIVE
will include old versions of sequence data sets.

   *** FOR VERSION RELEASE NOTES (below) SEARCH ON VERSION NUMBER. ***

===============================================================================


~/nuclear_cds_set.embl
======================

This sequence set was made by two methods. Sets up to and including
nuclear_cds_set.embl.v1.5 were made by method 1.  Set v2.0 and subsequent
where made by combining set v1.5 with that made by Method 2.

Method 1.
---------

        1. Records were retrieved from the EMBL and EMBLNEW data libraries at
        the EBI using the search terms "Drosophila melanogaster" + "CDS"
        with the public version of SRS5 (http://srs.ebi.ac.uk:5000).

        2. Using the AC numbers stored by FlyBase, FlyBase gene symbols were
        added to all records as:

        FT                   /db_xref="FLYBASE:<symbol>"

        3. The correspondence between AC number, FBgn number, PID and
        SWISS-PROT/TREMBL identifiers was checked against FlyBase. Some
        incorrect relationships (in both EMBL and FlyBase) were corrected
        for this dataset and were reported back to FlyBase (from where they
        will be reported to EBI/NCBI/DDBJ).

        4. For those records which lacked FlyBase identifier numbers
        the sequence AC & PID numbers were used to add both FlyBase gene
        symbols and FlyBase FBgn identifiers to all records. Records
        for genes not yet assigned FBgn identifiers have no

        FT                   /db_xref="FLYBASE:FBgn<integer>"

        line.

        5. All FlyBase genes encoded by the mitochondrial genome or
        transposons were filtered out.

        6. All FlyBase genes with >1 record were examined and filtered out
        so that for each gene there remains only a single record. The
        following filter criteria were used:

         a. If >1 record from the same gene and both start ATG and end
         with a terminator then only retain the longest. If >1 of the
         same length then choice was arbitary.

         b. All records start with ATG, unless a translation_exception
         is explicit feature qualifier in record (listed below). These all
         have an added:

         CC                   !!start_exception="<codon>"

         line.

         c. All records end with TAA|TAG|TGA. If this was missing from
         the feature then source record was examined in case of mis-
         -annotation. If not the record was deleted. If so, then record was
         edited and a comment line that begins:

         CC                   !!CDS annotation wrong <text>.

         added to the record. A list of these in included below.

         7. Records derived from Berkeley P1 sequences were filtered out.

         8. All extraneous FT and CC lines of source were deleted. Only
         these were retained:

         CC                   !!<text>
         FT   CDS             [nucleotides of parent sequence]
         FT                   /db_xref="FLYBASE:<symbol>"
         FT                   /db_xref="FLYBASE:FBgn<integer>"
         FT                   /db_xref="PID:<id>"
         FT                   /db_xref="SPTREMBL:<id>"
         FT                   /db_xref="SWISS-PROT:<id>"
         FT                   /translation="<translation>"

         9. All edited sequence records have this comment line:

         CC                   !!<text>

         All positions within these CC lines refer to daughter
         sequences, not parent, unless it is obvious from context
         that a direct comparison is being made.

Method 2.
---------

        1. A list of all genes and accessions in FlyBase were made available
        by Aubrey de Grey 12/15/98. Those already in nuclear_cds_set.embl.v1.5
        were removed, and the body of the remaining accessions were obtained
        from EMBL.

        2. Transposons, non-coding RNAs, pseudogenes, mitochondrial sequences
        and viruses were removed.

        3. All FlyBase genes with >1 record were examined and filtered out
        so that for each gene there remains only a single record. The
        following filter criteria were used:

         a. If >1 record from the same gene and both start ATG and end
         with a terminator then only retain the longest. If >1 of the
         same length then choice was arbitary.

         b. All full records start with ATG, unless a translation_exception
         is explicit feature qualifier in record (listed below). These all
         have an added:

         CC                   !!start_exception="<codon>"

         line.

         c. All records end with TAA|TAG|TGA. If this was missing from
         the feature then source record was examined in case of mis-
         -annotation. If so, then record was edited and a comment line
         -that begins:

         CC                   !!CDS annotation corrected from <coordinates>

         added to the record.

         4. a. For incomplete CDS's, the longest available CDS was selected.
         If the choice was between a longer incomplete CDS and shorter complete
         CDS, the shorter complete CDS was selected. If the choice was
         between a longer predicted CDS and a shorter experimental CDS, the
         shorter experimental CDS was selected.

         b. <> signs were entered around the coordinates of incomplete CDS's
         that had not been marked correctly in databank records. Subsequently,
         incomplete CDS's were automatically marked with:

         CC                   !!Thought to be an incomplete sequence.

         c. If translation of an incomplete CDS started at the 2nd or 3rd
         codon, the 5' orphan nucleotides were removed so all CDS's now
         start at with a complete codon. If there were only 2 nucleotides
         specifying the last codon, but if the codon assignment could be made
         unambiguously, these cases were marked with:

         CC                   !!Last base of last codon unknown, assignment
                               unambiguous.

         3' CDS sequences not assigned to a codon were removed.

         d. Fragments encoding less than 50 amino acids were removed.

         5. Predicted genes were marked with:
         CC                   !!Predicted gene.

         Records of EDGP and BDGP predicted genes were marked with:
         CC                   !!Predicted gene by EDGP.
         CC                   !!Predicted gene by BDGP.

         6. Genes predicted based on PCR fragments were marked with:

         !!Predicted based on PCR fragment.

         7. For either unannotated CDS entries, or for annotated CDS's
         that could be extended, new coordinates were curated by the FB
         curator, and marked with the CC line:

         CC                   !!CDS compiled by FB curator.

         If more than one accession was involved, the CC line includes
         the accession numbers:

         CC                   !!CDS compiled by FB curator from Acc#1 and Acc#2.


         8. The formatting of the file was done as a last step, to comply with
         step 8 of dros_sequence_sets.doc:

         All extraneous FT and CC lines of source were deleted. Only
         these were retained:

         CC                   !!<text>
         FT   source          <Accession_number:range>
         FT   CDS             <from..to>
         FT                   /db_xref="FLYBASE:<symbol>"
         FT                   /db_xref="FLYBASE:FBgn<integer>"
         FT                   /db_xref="PID:<id>"
         FT                   /db_xref="SPTREMBL:<id>"
         FT                   /db_xref="SWISS-PROT:<id>"
         FT                   /translation="<translation>"

         9. All edited sequence records have this comment line:

         CC                   !!<text>

         All positions within these CC lines refer to daughter
         sequences, not parent, unless it is obvious from context
         that a direct comparison is being made.


July 30 1997. v1.0:
====================

Database searched: EMBL version 51; SRS indexed 18 June 1997.
                   EMBLNEW indexed 5am 22 July 1997.

Source records:   2998.
Retained records: 1345.

Records edited:
  X04427; X75541; Y00133; X05245; X00854; X07311; X07181; X79243
  V00248; Y00049.

Non-ATG starts:
  U22825; M63724; Z14974; L11345.

Records deleted due to known problems:
  M25662.

Records compiled from >1 AC number:
  M25292+M25294 [given ID & AC XX1111]

Known problems:
  1. The cases of translation start correction need to be checked with
     literature. Cavener and Thummel asked for advice.

  2. In-frame stop codons, translated as seleno-cysteine, not dealt
     with. Only case known to be is kel and the 'second' 3' to TGA
     codon of L08483 is not included. Cooley asked for advice.

July 31 1997. v1.1:
====================

1. Cavener confirms the translational start exceptions as ok.

2. Cooley confirms in-frame TGA in kelch (FBgn0001301).

   ID   DMRCPA_4;  parent: DMRCPA
   and
   ID   DMRCPA_6;  parent: DMRCPA
   combined to:
   ID   DMRCPA_X;  parent: DMRCPA

   by joining both ORF's of L08483.

3. oaf (FBgn0011818) also has in-frame stop codon.

   ID   DMTNDPRO_5;  parent: DMTNDPRO
   and
   ID   DMTNDPRO_9;  parent: DMTNDPRO
   combined to:
   ID   DMTNDPRO_X;  parent: DMTNDPRO

   by joining both ORF's of L31349.

August 6 1997. v1.2:
====================

File cleaned as follows after checks by Martin Reese at LBL:

1. Mis-annotation of CDS of U61976 corrected (incorrect start).

2. Base 2034 of U49439 changed from 'S' to 'N'.

3. Following records deleted: Y09063; Y12701; K01664.

4. Following records edited to correct annotation of terminator: V00204; X04426;
   M62975; X15400; X58826;

August 13 1997. v1.3:
=====================

1. Following records deleted (partial): X01524; X16992; Y09064; Y10911; X78219;
   Y10910.

2. Following records edited to correct annotation of terminator: S41484; X03414;
   Y00308; X04896; X14215; X13331; X05245; V00204; M71249; X03121; X04426;
   Y00228; X07194.

3. The data set includes 1336 sequences.

August 25 1997. v1.4:
=====================

1. U73823 deleted (non-ATG start).

February 18 1999. v.1.5:
========================

1. This is the original training set, indentical in content to v1.4
   but with a correction to syntax of the 'source' and 'CDS' feature lines
   and updates on some of the db_xref lines.

March 1 1999. v2.0:
===================

1. This is the first released set that combines those made by Method 1
   and Method 2.  The reason for this was both to update the sequence
   sets and to provide better coverage for screening newly sequenced
   genomic data.  For the latter reason the second set (included in v2.0)
   includes partial sequences, and sequences annotated as predicted genes
   from the EDGP and BDGP sequence data.

   The following notes are in addition to those inherited from v1.5.

   Annotation corrected: Osbp (Y13951), Fs(2)Ket (AJ002729), byn(S74163).

   Annotation extended to include stop codon: dnt (AJ224361),
   Gyc32E (X72800).

   Complete CDS annotated: crm (Y13674),faf(L04959), Fdxh (X06542).

   Partial CDS annotated: kz (X64418), Rtc1(X04754), bcn92 (Z46608),
   gdl(X58286).

   CDS extended: lab (X13104,X13103), nAcR&bgr;-96A (Y14678,X55676),
   Acp98AB (U85762;U90949).

2. In this version the following genes are annotated as having non-ATG start
   codons:

   amn (FBgn0000076) (CTG)
   Cha (FBgn0000303) (GTG)
   cpo (FBgn0000363) (CTC)
   ewg (FBgn0005427) (CTG)
   l(2)k10201 (FBgn0016970) (CTG)
   anon-fe2D7 (FBgn0022342) (CTG)

3. In this version the following genes are annotated as having in-frame
   termination codons:

   oaf (FBgn0011818) (translated as X)
   kel (FBgn0001301) (translated as X)

3. This combined set includes 2053 sequences.

March 17 1999. v2.1:
====================

1.  Addition of new records; total number 2062.

April 15 1999. v2.2:
====================

1. A parse error which resulted in the retention of old source & CDS
   lines towards the end of the file (from the record for &agr;Cop to
   end of file) has been corrected. A glitch in the annotation of the
   Eaat1 record has been corrected.

2. Only records new (or changed) in this release have the new protein_id from
   the nucleic acid sequence database.  The older records still have
   PIDs.  This will be fixed for the next release.

3. The number of records is now 2070.

May 24 1999. v2.3:
==================

1. The PID to protein_id translation is still not complete.

2. Corrections (updates) have been made to gene symbols, following
   changes in FlyBase.  Three records (FBgn0002354, FBgn000342,FBgn0004514)
   have been replaced by longer sequences.

3. Addition of new records; total number 2122.

July 31 1999. v2.6:
==================

0. versions v2.4 and v2.5 were internal and not released.

1. All PIDs are now translated to the corresponding protein_ids.

2. Corrections (updates) have been made to gene symbols, following
   changes in FlyBase.  Records for the following genes have been
   replaced by new records: Ddc, Dref, Prat, sisA, LanB1, Egfr, EG:56G7.1.

3. Records for 18 genes have been removed due to the fact that they
   were duplicates (due to having had different gene symbols in the past).

4. This data set now includes short CDSs (<150-bp).

5. The number of records is: 2193

October 3 1999. v2.7:
=====================

1. A number of FlyBase symbols have been updated.

2. As a consequence of an ongoing collaboration with Eleanor Whitfield
   of the EBI SwissProt team a number of EMBL/GenBank/DDBJ records have
   now updated FT lines, resulting in the removal of several !!CC lines
   from these records.  In addition some of the translations that had
   been previously compiled by FlyBase are now in the nucleic acid
   sequence records.  The records for the following genes are affected:
   kz, bcn92, Rtc1, Gyc32E, Fdxh, gdl.

3. Records for the following genes have been replaced by newer versions:
   crm, Hr38, lok, mus201, oaf, Pfk, Pu, Est-P.

4. Two duplicate records (BcDNA:LD24492, BcDNA:GM07659) have been removed.

5. The CTC "start exception" for the cpo gene was an error; this has been
   corrected.  The interpretation of the in-frame stop codon within the oaf
   gene has changed; only the first ORF of this gene is now included.

6. This data set includes translations from genes predicted by
   the Drosophila genome projects, including the BDGP cDNAs.  It does
   not include translations of BDGP EST sequences.

7. The number of records is: 2452.

November 8 1999. v2.8.5:
========================

1.  The number of records is now 2546.

2.  Apart from routine updates and corrections to symbols the major
    change initiated with this release is to coordinate this sequence
    set with the curation of gene structures by FlyBase.  These records
    have FBpp db_xref identifier numbers.

January 3 2000. v2.9:
=====================

1.  The number of records is now 2636.

2.  Six sequences have been removed (Bkm, Ptth, Tk3, Tk4, Tk6, Sry-rDM17) as we
    have good reasons to believe that they were not Drosophila.

3.  The following duplicate records have been removed: BcDNA:GM12270, dek,
    Ac35C, BcDNA:GM07659, BcDNA:LD07532, BcDNA:LD20207, BcDNA:LD21772,
    BcDNA:LD24440 and ubiquitin-C-terminal-hydrolase.

4.  A number of symbols have been updated.

5.  The protein_id's of the records for aub and EG:196F3.2 have been
    corrected.

6.  The sequences for the following genes have been replaced with longer
    or different versions: Fak56D, Fcp3C, N, prominin-like, Pten, Rab4, rdgA
    RecQ5.



===============================================================================

~/unique-50-genes.80.fa
~/unique-50-genes.50.fa
=======================

These data sets were made from the ~/nuclear_cds_set.embl file
by Martin Reese at LBL, Berkeley. They have been obtained from:

ftp://www-hgc.lbl.gov/pub/genesets/dro/unique-50-genes.fa
ftp://www-hgc.lbl.gov/pub/genesets/dro/unique-80-genes.fa

BLASTN was used in an all-against-all comparison of the nucleic acid
sequences. From any clump of sequences with an overall identity above 80%
(or 50%) only one member was chosen (at random) for inclusion.

August 13 1997. v1.3:
=====================

Derived from nuclear_cds_set.embl.v1.3.

The 80% cutoff data set includes 1114 sequences; the 50% cutoff set, 866.

===============================================================================

mitochondrial_cds_set.embl
==========================

This sequence set includes the coding sequences of mitochondrial
proteins.  They are derived from the EMBL accessions U37541 and M37275.
The protein sequences are translated with Translation table 5
of the International Nucleotide Sequence Data Library
(http://www.ebi.ac.uk/ebi_docs/embl_db/ft/genetic_codes.html).
Coding sequences with start exceptions are noted in CC lines, e.g.:

CC                   !!Translation_start_exception="ATAA".

and non ATG starts are noted, e.g.:

CC                   !!Translation_start="ATA".

January 30 1999. v1.0:
======================

This is the first released set and includes all thirteen known
sequences of the mitochondrial genome that encode proteins.

===============================================================================

~/repeat_sequence_set.embl
==============================



Method.  Records for 'gene' objects annotated in FlyBase as representing a
         'repetitive sequence' or 'repetitive element' (in *v field) were
         retrieved from the EMBL and EMBLNEW data libraries at the EBI using
         the SRS5 command line interface (getz). Records corresponding to
         transposable elements of known structural classes were filtered.
         Using the FT descriptions of the sequences a script was used to trim
         extraneous sequences, leaving only that annotated as being repetitive.
         Identical sequences were removed by a filter, but no attempt has been
         made to reduce redundancy in this data set.

         The sequence retained from the original record is described in a
         line with this syntax:

         FT   source         <AC>:<first_base>..<last_base>

         where AC is the primary accession number of the parent record.

         The records retain the original ID lines of their EMBL record
         parents, with the substitution of the length of the derived sequence.
         They retain the DR (data base cross_reference) lines pointing
         to the corresponding FlyBase gene identifier and symbol and their
         parent's AC and NI lines.

         The criteria for inclusion in this data set do not include
         repetitive coding sequences. This data set excludes, therefore,
         sequences coding for homopolymeric amino-acid runs (e.g. the opa
         and pen repeats) and motifs such as the EGF-repeat.

August 19 1997. v1.0
====================

1. Consists of 89 records; total length 46,489-bp.

2. Includes entries from 24 different FlyBase gene objects.

FBgn0004084    &agr;&ggr;-element.
FBgn0004950    1688-3C.
FBgn0004951    1688-10Ea.
FBgn0004952    1688-10Eb.
FBgn0005587    anon-alaA.
FBgn0005588    anon-alaE.
FBgn0005589    anon-alaD.
FBgn0015362    anon-k.
FBgn0011748    anon-M/SAR.
FBgn0014123    anon-Xh1.
FBgn0000192    Bkm.
FBgn0004633    CAT-repeat.
FBgn0005663    dodeca.
FBgn0004142    GATA-sequence.
FBgn0004409    het-rep1.
FBgn0004411    het-rep3.
FBgn0005668    HMR-element.
FBgn0015786    Porto1.
FBgn0003315    satDNA.
FBgn0003561    suffix.
FBgn0003582    Su(Ste).
FBgn0015563    vivi-repeat.
FBgn0004081    XDm.
FBgn0004083    Y-seqs.

August 23 1997. v2.1
====================

1. Consists of 96 records; total length 64,075-bp.

2. The following entries have been added following analysis of Berkeley na_gb.trans set:

FBgn0002643    mam. [RS repeat]
FBgn0000164    AluI-like repeat from rDNA.
-              His3-His1 spacer sequence.
FBgn0000126    Ars4.
FBgn0000128    Ars6.
FBgn0000002    5SRNA. [Repeat unit.]
FBgn0000164    bb. [Complete rDNA repeat unit.]
FBgn0003284    Rsp.
-              Repetitive sequence flanking su(f).
FBgn0005668    HMR-element. [#2]

3. These sequences have been removed:

FBgn0005587    anon-alaA.
FBgn0005588    anon-alaE.
FBgn0005589    anon-alaD.

===============================================================================

~/transposon_sequence_set.embl
==============================

Method.  Records for 'gene' objects annotated in FlyBase as representing
         transposon (in the *t field) were retrieved from the EMBL and
         EMBLNEW data libraries at the EBI using the SRS5 command line.
         For each transposon all records were screened by eye to select
         the longest sequence not noted as being from a defective element.
         In all possible cases the annotation in the FT lines was checked
         against that published in the literature; in case of conflict the
         literature was preferred over the sequence annotation. Extraneous
         sequences, including target site duplications, were
         trimmed and joins were made across records in an attempt to
         assemble a 'virtual' sequence.

         In this data set there is only a single record for each
         transposon. Many, if not all, transposons exist in variant
         forms and no attempt has been made to annotate these differences.

         Only three features have been carried across from the parent
         sequence records - these are the limits of the direct or
         inverted repeat ends of the element (if present), the open reading
         frames (CDS) and the introns.

         The sequence retained from the original record is described in a
         line with this syntax:

         FT   source         <AC>:<first_base>..<last_base>

         Joins are annotated in an 'FT   source' line with this syntax:

         join(<AC_1>:<first_base>..<last_base>,<AC_2><first_base>..<last_base>)

         where AC_1 and AC_2 are the primary accession numbers of the
         parent sequences.

         Sequences that are the complement of their parent are annotated
         with this sytax:

         complement(<AC>:<first_base>..<last_base>)

         The records retain the original ID lines of their EMBL record
         parents, with the substitution of the length of the derived sequence.
         They retain the DR (data base cross_reference) lines pointing
         to the corresponding FlyBase gene identifier and symbol and their
         parent's AC and NI lines.

         The Drosophila genome sequencing projects will identify both
         new elements and more 'complete' sequences of known elements.
         These sequences will be used in this data set.

Annotation: These sequences have minimal annotation; only open reading
         frames, introns and the extents of the terminal repeats (if present)
         are annotated. This annotation derives from the original sequence
         record but has, in all cases, been checked, and if necessary,
         supplemented, against the original publication and by analysis
         of the sequences.

August 19 1997. v1.0
====================

The data set consists of records from 40 different transposons. These
records have no annotation of CDS or terminal repeat features:

FB gene ID     Symbol            EMBL              Size       Comment

Retroviral elements:

FBgn0000004    17.6              X01472            7439bp     complete
FBgn0000007    1731              X07656            4648bp     complete
FBgn0000005    297               X03431            6995bp     complete
FBgn0005384    3S18              U23420            6126bp     complete
FBgn0000006    412               X04132            6897bp     complete
FBgn0010103    aurora-element    X70361             344bp     incomplete; LTR
FBgn0014947    blastopia         Z27119            5034bp     complete
FBgn0000199    blood             AY180916          7419bp     complete
FBgn0010302    Burdock           U89994            6411bp     complete
FBgn0000349    copia             D90356            5143bp     complete
FBgn0000481    Doc               X17551            4725bp     ?incomplete
FBgn0015945    gate              X97139             245bp     incomplete; LTR
FBgn0001167    gypsy             M12927            7469bp     complete
FBgn0001207    HMS-Beagle        J01078             266bp     incomplete; LTR
FBgn0002697    mdg1              X59545            7480bp     complete
FBgn0002698    mdg3              X95908            5519bp     complete
FBgn0002745    micropia          X14037            5457bp     complete
FBgn0000155    roo               Z48503            2138bp     incomplete; cDNA
FBgn0003490    springer          M10908             405bp     incomplete; LTR
FBgn0003519    Stalker           X86075            1651bp     incomplete
FBgn0004082    Tirant            X93507            2484bp     incomplete

LINE-like retrotransposons:

FBgn0000224    BS                X77571            5126bp     ?complete
FBgn0000430    D-element         X05643             408bp     incomplete
FBgn0000652    F-element         M17214            3520bp     incomplete
FBgn0001100    G-element         X06950            4346bp     ?complete G3A
FBgn0004141    HeT-A             U06920            6943bp     complete
FBgn0001249    I-element         M14954            5371bp     complete
FBgn0001283    jockey            M22874            5020bp     complete
FBgn0003908    R1-element        X51968            5356bp     complete
FBgn0003909    R2-element        X51967            3607bp     complete
FBgn0004904    TART-element      U14101           10654bp     ?complete

IR-elements:

FBgn0005773    Bari1             X67681            1728bp     complete
FBgn0001210    hobo              M69216            2959bp     complete
FBgn0001181    HB                X01748            1653bp     ?incomplete
FBgn0014967    hopper            X80025            1435bp     incomplete
FBgn0003055    P-element         X06779            2907bp     complete
FBgn0003122    pogo              X59837            2121bp     ?incomplete
FBgn0004905    S-element         U33463            1736bp     ?incomplete
FBgn0002949    NOF               X15469;X51937     4347bp     complete

FB-elements:

FBgn0000638    FB                V00246            1492bp     ?incomplete

August 25 1997. v2.0
====================

1. The sequences have been annotated for ORFs, introns and terminal repeats.
   These features have been checked against the original publications, except
   for X95908, but have not yet been checked by sequence analysis.

2. The sequence trimming of the following records has been corrected:
   join(X15469;X51937); X86075.

October 07 1997. v3.0
=====================

1. The roo sequence (Z48503) has been replaced by a 9092-bp sequence
   from a Berkeley P1 clone.

2. A new retroviral element, yoyo, similar to the element of the same
   name from the Medfly, has been added from a Berkeley P1 clone.

3. The sequence of the 1360 element, inadvertently missed in previous
   versions, has been added.

October 10 1997. v3.1
=====================

1. The Tirant sequence has been updated to that of the complete element.

November 24 1997. v3.2
======================

1. The blood sequence has been updated to that of the complete element,
   from a Berkeley P1 clone sequence.

2. The FlyBase gene identifier has been added to the yoyo record.

November 24 1997. v3.3
======================

1. The ZAM element has been added.

September 8 1998. v3.4
======================

1. The full length sequence of the GATE element has been added.

2. The FlyBase gene identifier has been added to the ZAM record.


February 1 1999. v3.6
=====================

1. The full length sequence of the Idefix element has been added.

March 27 1999. v3.7
===================

1. A more complete sequence of the aurora element has been added,
   replacing the LTR sequence previously included.

May 6 2000. v3.8
================

1. K. O'Hare, Personal communication to FlyBase, 1 May 2000:
   The pogo sequence is probably complete element.

2. The sequence of the F element has been replaced by one of 4708bp
   from K. O'Hare (Personal communication to FlyBase, 1 May 2000).

3. The following three new elements have been added: X-element, Transpac.
   and Circe.

November 26 2000. v.4.0
=======================

1. You element added.

November 26 2000. v.4.1
=======================

Line 190 of the file had, mysteriously, lost a base. Now corrected.
Thankyou Suzi for discovering this !

April 9 2001. v.4.2
===================

1. Full length springer and cruiser elements added from BDGP.

2. midline element added.


April 9 2001. v.4.3/4.31
========================

1. Full length HMS Beagle element added from BDGP.

2. In v4.31 there have been updates to FBgn and AC numbers; If these are
   not known then 'n' filled dummies are used.  The NI lines,
   sometimes containing the NCBI internal tracking number, and the FH
   lines (which contained no information) have been deleted.

May 4 2001. v.4.4
=================

1. The BEL-like Tinker element sequence has been added.

May 4 2001. v.4.5
=================

1. Updated AC etc data on springer, HMS-Beage and cruiser.

July 29 2001. v.4.5
===================

1. Added FBgn numbers for Tinker & cruiser; FBgn for yoyo corrected.

August 21 2001. v.4.6
=====================

1. midline element sequence deleted; it is an internally deleted
form of HMS Beagle.

September 21 2001. v.4.7
========================

1. The sequence of the pilgrim element has been added.

2. The incomplete sequence of the Stalker element (EMBL:X86075) has been
   replaced by a full-length sequence from the BDGP.

September 21 2001. v.4.71
=========================

1. AC lines with no accession number, i.e.
   AC   nnnnnnnnn;
   have been removed.

2. Some extraneous text was appended to the end of the Stalker sequence;
   Now removed.


September 21 2001. v.4.73
=========================

1. Sequence of INE-1 element added.

November 21 2001. v.4.8
=======================

1. The sequences of the new elements identified by REPBASE
   http://www.girinst.org/server/RepBase/RepBase6.6.embl/drorep.ref
   have been assembled by Josh Kaminker and added.

2. The sequences of the Tirant and BS elements have been replaced with
   sequences from the BDGP.

November 26 2001. v.4.82
========================

1. REPBASE copia2 sequence added.

November 26 2001. v.4.9
=======================

1. Following the identification of synonyms the following sequences have been
   removed from the set:

   Ivk (same element as You)
   Waldo-B (same element as pilger)
   diver (same element as Tinker)
   tabor (same element as Pilgrim)
   D-element (same element as jockey)

December 20 2001. v.4.91
========================

1. The sequence of the strider element has been added.

January 26 2002. v.4.92
=======================

1. The Het-A sequence has been revised.

2. The sequences of Juan, forgger and rover have been added.

February 12 2002. v.4.93
========================

1. The Rt1b sequence has been removed; it is the same element as Waldo-A.

March 13 2002. v.4.94
=====================

1. The strider sequence has been removed; it is the same element as Juan.

2. The M55078 sequence of 1360 is replaced by the REPBASE "protop"
   sequence; but the valid name of this element remains 1360.

3. The FlyBase names of the following elements have been changed:
   cruiser to Quasimodo
   Pilgrim to Tabor
   Tinker to diver
   You to Ivk
   Waldo-A to Rt1b

March 15 2002. v.4.95
=====================

1. Name of yoyo changed to opus.

March 15 2002. v.4.96
=====================

1. Duplicate HMS-Beagle record (J01078) removed.

May 4 2002. v.4.97
==================

1. Name update.

August 15 2002. v.5.00
======================

1. Name & FBgn updates.

2. Following added from Repbase 4.4.3: jockey2, looper1, Tom1, G4, G5, G6.

3. FB element corrected to remove 3' HB element sequences.

4. A duplicated record of transib2 has been removed.

August 15 2002. v.5.01
======================

1. Partial Penelope sequence added.

August 28 2002. v.5.02
======================

1. The following sequences have been added from the BDGP: qbert, McClintock, 
   hopper2, Stalker4.

November 14 2002. v.6.2
=======================

1. Sequence of Bari2, Max & Stalker2 elements added.

2. FBid's & AC numbers updated.

3. narep1 sequence deleted (synonym of INE-1).

November 14 2002. v.7.1
=======================

1. The sequence of 412 was found to be deleted. A new sequence
   from Release 3 has replaced it.

2. CDS translations (from SWALL) have been added to many of the records.
   These have not yet been checked against other annotation.

December 28 2002. v.7.2
=======================

1. Updates of FB identifiers and additions of CDS translations of blood, roo,
   opus and Juan.

October 25 2003. v.7.2.1
=========================

1. Updates of FB identifiers.

The current data set is:

FB gene ID     Symbol            EMBL              Size       Comment

Retroviral elements:

FBgn0000004    17.6              X01472            7439bp     complete
FBgn0000007    1731              X07656            4648bp     complete
FBgn0000005    297               X03431            6995bp     complete
FBgn0005384    3S18              U23420            6126bp     complete
FBgn0000006    412               nnnnnnnn          7567bp     complete
FBgn0063447    accord            nnnnnnnn          7404bp     complete
FBgn0010103    aurora-element    X70361            4263bp     ?complete
FBgn0014947    blastopia         Z27119            5034bp     complete
FBgn0000199    blood             nnnnnnnn          7410bp     complete
FBgn0010302    Burdock           U89994            6411bp     complete
FBgn0022937    Circe             X98424            6356bp     complete
FBgn0000349    copia             D90356            5143bp     complete
FBgn0062343    Dm88              nnnnnnnn          4558bp     complete
FBgn0044355    Quasimodo         AF364550          7387bp     complete
FBgn0063439    diver2            nnnnnnnn          4917bp     complete
FBgn0061513    frogger           AF492763          2483bp     ?complete
FBgn0015945    GATE              AJ010298          8507bp     complete
FBgn0063436    gtwin             nnnnnnnn          7411bp     complete
FBgn0001167    gypsy             M12927            7469bp     complete
FBgn0063435    gypsy2            nnnnnnnn          6841bp     complete
FBgn0063434    gypsy3            nnnnnnnn          6973bp     complete
FBgn0063433    gypsy4            nnnnnnnn          7369bp     complete
FBgn0063432    gypsy5            nnnnnnnn          6852bp     complete
FBgn0063431    gypsy6            nnnnnnnn          7826bp     complete
FBgn0001207    HMS-Beagle        AF365402          7062bp     complete
FBgn0026065    Idefix            AJ009736          7411bp     complete
FBgn0063430    invader1          nnnnnnnn          4032bp     complete
FBgn0063429    invader2          nnnnnnnn          5124bp     complete
FBgn0063428    invader3          nnnnnnnn          5484bp     complete
FBgn0063427    invader4          nnnnnnnn          3105bp     complete
FBgn0063426    invader5          nnnnnnnn          4038bp     complete
FBgn0063919    Max-element       AJ487856          8556bp     complete
FBgn0063917    McClintock        AF541948          6450bp     complete
FBgn0002697    mdg1              X59545            7480bp     complete
FBgn0002698    mdg3              X95908            5519bp     complete
FBgn0002745    micropia          X14037            5457bp     complete
FBgn0063782    qbert             AF541947          7650bp     complete
FBgn0045970    Tabor             nnnnnnnn          7345bp     complete
FBgn0063450    Tom1              nnnnnnnn           410bp     incomplete
FBgn0000155    roo               AY180917          9092bp     complete
FBgn0063394    rooA              nnnnnnnn          7621bp     complete
FBgn0061485    rover             AF492764          7318bp     complete
FBgn0003490    springer          AF364549          7546bp     complete
FBgn0003519    Stalker           AF420242          7365bp     complete
FBgn0063455    Stalker2          nnnnnnnn          7672bp     complete
FBgn0063454    Stalker3          nnnnnnnn           372bp     incomplete
FBgn0063897    Stalker4          AF541949          7359bp     complete
FBgn0043969    diver             nnnnnnnn          6112bp     complete
FBgn0040267    Transpac          AF222049          5249bp     complete
FBgn0004082    Tirant            nnnnnnnn          8526bp     complete
FBgn0003007    opus              AY180918          7521bp     complete
FBgn0023131    ZAM               AJ000387          8435bp     complete

LINE-like retrotransposons:

FBgn0063440    baggins           nnnnnnnn          5453bp     complete
FBgn0000224    BS                nnnnnnnn          5142bp     complete
FBgn0063594    Cr1a              nnnnnnnn          4470bp     complete
FBgn0000481    Doc               X17551            4725bp     ?incomplete
FBgn0063534    Doc2-element      nnnnnnnn          4789bp     complete
FBgn0063533    Doc3-element      nnnnnnnn          4740bp     complete
FBgn0000652    F-element         AC005198          4708bp     complete
FBgn0001100    G-element         X06950            4346bp     ?complete G3A
FBgn0063507    G2                nnnnnnnn          3102bp     complete
FBgn0063506    G3                nnnnnnnn          4605bp     complete
FBgn0063505    G4                nnnnnnnn          3856bp     complete
FBgn0063504    G5                nnnnnnnn          4856bp     complete
FBgn0063503    G6                nnnnnnnn          2042bp     complete
FBgn0004141    HeT-A             U06920            6083bp     complete
FBgn0001249    I-element         M14954            5371bp     complete
FBgn0043055    Ivk               nnnnnnnn          5402bp     complete
FBgn0046110    Juan              AY180919          4236bp     complete
FBgn0001283    jockey            M22874            5020bp     complete
FBgn0063425    jockey2           nnnnnnnn          3428bp     complete
FBgn0046701    Penelope          AF418572           804bp     incomplete
FBgn0003908    R1-element        X51968            5356bp     complete
FBgn0003909    R2-element        X51967            3607bp     complete
FBgn0041728    Rt1a              AJ278684          5108bp     complete
FBgn0063467    Rt1c              nnnnnnnn          5443bp     complete
FBgn0042682    Rt1b              AF281636          5171bp     complete
FBgn0004904    TART-element      U14101           10654bp     ?complete
FBgn0042231    X-element         AF237761          4740bp     complete

IR-elements:

FBgn0005673    1360              nnnnnnnn          3409bp     complete
FBgn0005773    Bari1             X67681            1728bp     complete
FBgn0064134    Bari2             AF541951          1064bp     complete
FBgn0001210    H-element         M69216            2959bp     complete
FBgn0001181    HB                X01748            1653bp     ?incomplete
FBgn0014967    hopper            X80025            1435bp     incomplete
FBgnnnnnnnn    hopper2           AF541950          1593bp     incomplete
FBgn0026416    INE-1             U66884             611bp     ?incomplete
FBgn0063402    looper1           nnnnnnnn          1881bp     incomplete
FBgn0063401    mariner2          nnnnnnnn           912bp     complete
FBgn0002949    NOF               X15469;X51937     4347bp     complete
FBgn0003055    P-element         X06779            2907bp     complete
FBgn0003122    pogo              X59837            2121bp     complete
FBgn0004905    S-element         U33463            1736bp     ?incomplete
FBgn0063466    S2                nnnnnnnn          1735bp     complete
FBgn0026410    Tc1               nnnnnnnn          1666bp     complete
FBgn0063372    transib1          nnnnnnnn          2167bp     complete
FBgn0063371    transib2          nnnnnnnn          2844bp     complete
FBgn0063370    transib3          nnnnnnnn          2883bp     complete
FBgn0063369    transib4          nnnnnnnn          2656bp     complete

Foldback elements:

FBgn0000638    FB                V00246            1106bp     ?incomplete

===============================================================================