Skip to content
Snippets Groups Projects
Commit b802ac29 authored by Graham McVicker's avatar Graham McVicker
Browse files

ideas about haplotypes

parent e5bd5d16
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
docs/ensembl_changes_spec.txt
+ 166
− 22
View file @ b802ac29
ENSEMBL - API Change Specification
==================================
INTRODUCTION
------------
This document is intended as an outline of possible changes to the current
incarnation of the EnsEMBL database schema and Perl API. This document is
evolving and will hopefully enable us to develop a plan for the progression
of the EnsEMBL API. Nothing is finalized and everything is open to change.
Many of the proposed alterations may need to be phased in over time or may
never occur at all.
REVISION HISTORY
----------------
Graham Mcvicker - July 9, 2003 - Created
Graham Mcvicker - July 9, 2003 - Created
Graham McVicker - July 22, 2003 - Added Haplotype Spec.
GOALS
-----
-A cleaner, more intuitive API
-A more general schema able to better capture divergent assembly types
-More flexibility with regards to assembly related data such as haplotypes
MHC regions etc.
-More flexibility with regards to assembly related data such as haplotypes,
PARs, WGS assemblies etc.
SCHEMA MODIFICATIONS
--------------------
......@@ -113,7 +125,8 @@ Removed Tables
prediction_transcript
---------------------
PTs are now stored as ordinary transcripts without genes. They should
probably be computed in chromosomal coordinates instead of contig coordinates.
probably be computed in chromosomal coordinates instead of contig
coordinates.
contig
......@@ -207,8 +220,8 @@ Slice
The concept of an 'empty' slice will no longer exist.
The get_tiling_path method will have to be implemented differently for the
new system. It will take a coord_system string as an argument and rather than
returning a list of Tile objects it will return a listref of triplets
new system. It will take a coord_system string as an argument and rather
than returning a list of Tile objects it will return a listref of triplets
containing a start int, and end int, and a 'to' slice object. The following
is an example of how this method would be used ($clone is a reference to a
slice object in the clone coordinate system):
......@@ -234,14 +247,20 @@ SliceAdaptor
system, frag_name, start, end, and strand. The old, commonly used method
fetch_by_chr_start_end can be altered to simply chain to this new method
as can most other SliceAdaptor methods.
fetch_by_coord_system_frag_start_end_strand ?
Another method which will be necessary with the disapearence of the Clone,
RawContig and Chromosome adaptors is the one which allows for all slices
of a certain type to be retrieved. For example it is often necessary to
retrieve all chromosomes, or clones for a species. This method could be
named fetch_all_by_coord_system or something similar.
Proposed Method Names and Signatures
------------------------------------
Slice fetch_by_region(coord_system, name)
Slice fetch_by_region(coord_system, name, start)
Slice fetch_by_region(coord_system, name, start, end)
Slice fetch_by_region(coord_system, name, start, end, strand)
listref of Slices fetch_all(coord_system)
RawContig
---------
......@@ -370,7 +389,7 @@ Root
Storable Base Class
-------------------
Almost all business objects in the EnsEMBL system are storable in the database
Almost all business objects in the EnsEMBL system are storable in the db
and the ones which are always require 2 methods: dbID and adaptor. It would
make sense in order to reduce code duplication to have all storable objects
inherit from a Storable base class which implemented these.
......@@ -543,9 +562,9 @@ FeatureAdaptors
Bio::Seq
--------
Bio::Seq objects are not useful and inconsistant with some of the API methods.
Some methods return a string while others return a Bio::Seq. In the
instances where a Bio::Seq is returned the most common usage is to simply
Bio::Seq objects are not useful and inconsistant with some of the API
methods. Some methods return a string while others return a Bio::Seq. In
the instances where a Bio::Seq is returned the most common usage is to simply
extract the sequence with a call like: $seq = $obj->seq->seq; For
consistancy Bio::Seq usage will be removed in favour of simple string usage.
*Note: This change can be done independently of the other changes. It may be
......@@ -554,11 +573,143 @@ Bio::Seq
once)
Haplotypes
----------
TBD
NEW FEATURES
------------
Haplotypes (and the MHC region)
-------------------------------
There are several requirements related to haplotypes:
- Must be able to determine which haplotypes overlap a slice
- Must be able to run genebuild/raw computes over the haplotypes
- Must be able to retrieve a slice on a haplotype and its flanking
regions (i.e. the regions of the default assembly bordering the
haplotype).
Proposal:
Store haplotypes as features in a haplotype feature table, but also as
a single entry in the dnafrag table. The dnafrag table will contain an
extra column 'flags' of type set which will allow additional properties
to be associated with a dnafrag. If the haplotype dnafrag is assembled
from smaller dnafrags then this will be described by the assembly table
as usual (actually it may be necessary to do this if dna is not associated
directly with the haplotype coordinate system).
The haplotype table will look similar to the assembly table, but can be
viewed as a feature table that describes haplotypes:
haplo_feature
-------------
haplo_feature_id int
dnafrag_id int
dnafrag_start int
dnafrag_end int
dnafrag_strand int (may not be needed, may implicitly be 1)
haplo_dnafrag_id int (this references the dnafrag that is a haplotype)
With the addition of a HaplotypeFeatureAdaptor it will be possible to
retrieve haplotypes which overlap a slice. A HaplotypeFeature will inherit
from the Feature class and will have the basic 'start', 'end', 'strand',
'slice', attributes, but also a 'name' and 'length' retrieved from the
associated dnafrag.
It is possible to retrieve a slice on a haplotype just as any other slice
is retrieved from the SliceAdaptor. For example:
$slice = $slice_adaptor->fetch_by_region('chromosome', '6_DR52');
The slice will have an additional method is_type('haplotype') which
will return true or false depending on whether the slice is a haplotype
slice or not. This method will have to query the dnfrag table and
cache the result.
A slice created on a haplotype will have coordinates relative to the
start of the haplotype region NOT the start of the entire chromosome
as a normal slice. Furthermore creating a slice that extends past the
the boundaries of the slice (i.e. start < 1 or end > length) will
allow the user to retrieve flanking features and sequence from the
default region. The feature adaptors fetch_by_Slice method will have to
be altered to check whether a slice is a haplotype slice. If it is a
haplotype slice then the following algorithm will apply:
(a) Split the slice into 3 slices on the following regions:
(1) slice_start -> 0,
(2) 1 -> frag_length,
(3) frag_length + 1 -> slice_end
Slices that would be of length < 1 are not created or used.
Slices (1) and (3) are normal non-haplotype slices that are created
on the default chromosome which will be determined by querying the
haplo_feature table. Slice (2) is a haplotype slice.
The case in which it is not necessary to create slices (1) and (3) is
the degenerative case: it is not necessary to create any new slices
and the features for this slice can be returned right away.
(b) Retrieve features by recursively calling fetch_by_Slice on each of
the slices created.
(c) Adjust the start/end of features from slice (2) by adding the
the length of slice (1).
(d) Adjust the start/end of features from slice (3) by adding the
combined lengths of slice (1) and slice (2)
(e) Return all of the features retrieved
The SequenceAdaptor will also need to take haplotype slices into account.
A similar algorithm to above will need to apply in which up to 3 seperate
slices are created, seperate sequence for all three obtained and the
sequence from each of them obtained, spliced together and returned.
Pseudo Autosomal Regions (PARs)
-------------------------------
TBD
Multiple Assemblies
-------------------
TBD
Circular Chromosomes
--------------------
We can handle circular chromosomes (or any arbitrary circular sequence) in
a similar way the the haplotypes. The dnafrag for the circular sequence can
have a flag set which indicates that it is circular. The slice would have
an additional method is_type('circular') which would return true if the
slice was on a circular dnafrag. The following is the algorithm for
retrieval of features on a circular slice:
(a) Split the slice into 3 regions:
(1) slice_start -> 0,
(2) 1 -> frag_length,
(3) frag_length + 1 -> slice_end
(b) Create slices on each of the regions.
Region (1) becomes a circular slice with:
start = frag_length - region_length + 1
end = frag_length
Region (2) just creates a circular slice of that region
Region (3) becomes a circular slice with:
start = 1 - region_length
end = 0
Slices that would be of length < 1 are not created or used.
The case in which it is not necessary to create slices (1) and (3)
is the degenerative case: it is not necessary to create any new
slices and the features for this slice can be returned right away.
(c) Retrieve features by recursively calling fetch_by_Slice on each of
the slices created.
(d) Adjust the start/end of features from slice (2) by adding the
the length of slice (1).
(e) Adjust the start/end of features from slice (3) by adding the
combined lengths of slice (1) and slice (2)
(f) Return all of the features retrieved
A very similar algorithm would be applied with regards to sequence retrieval.
Comparative Sequence (Chimp)
----------------------------
TBD
OTHER CONSIDERATIONS
......@@ -573,10 +724,3 @@ Feature Transfer Accross Assemblies
change. We will supply a mechanism via which features from a previous
assembly may be transfered to a new assembly.
TBD
MHC Regions
Circular Chromosomes
Multiple Assemblies
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment