Re: nd_assay_genotype

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: nd_assay_genotype

seth redmond
whilst loading in some test data I encountered a number of cases where  
two genotypes will arise from a single assay (best example: a single  
staining will produce a karyotype for multiple chromosomes). As a  
result, for my DB, I've altered the restriction on nd_assay_genotype  
from UNIQUE(genotype_id) to UNIQUE(genotype_id, nd_assay_id).

I can't see any compatibility issues with the old schema, so does  
anyone have any objections if I roll this into the main source?

-s




--
Seth Redmond
   Scientific Programmer, VectorBase
   Kafatos / Christophides Groups
   Div. Cell and Molecular Biology
   Imperial College, London
[hidden email]
--


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Kitsos Louis

Hi Seth,

What does this really mean?

(best example: a single staining will produce a karyotype for multiple chromosomes). 

And how do genotypes arise from assays?

K.

************************************************
Prof. Christos (Kitsos) Louis
IMBB-FORTH & Dept. Biol., U. Crete
N. Plastira 100
700 13 Heraklion, Crete, Greece
************************************************
tel office:   +30-281 0391 119 and
                 +30-281 0391 199 (after work)
tel mobile: +30-693 2735 566
fax:           +30-281 0391 104
************************************************


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Kitsos Louis
In reply to this post by seth redmond

Hi Seth,

What does this really mean?

(best example: a single staining will produce a karyotype for multiple chromosomes). 

And how do genotypes arise from assays?

K.

************************************************
Prof. Christos (Kitsos) Louis
IMBB-FORTH & Dept. Biol., U. Crete
N. Plastira 100
700 13 Heraklion, Crete, Greece
************************************************
tel office:   +30-281 0391 119 and
                 +30-281 0391 199 (after work)
tel mobile: +30-693 2735 566
fax:           +30-281 0391 104
************************************************


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd_assay_genotype

Sook Jung
In reply to this post by seth redmond
Hi Seth
No objections - I think we need that change.
Sook

On Wed, Jul 14, 2010 at 1:09 PM, seth redmond
<[hidden email]> wrote:

> whilst loading in some test data I encountered a number of cases where
> two genotypes will arise from a single assay (best example: a single
> staining will produce a karyotype for multiple chromosomes). As a
> result, for my DB, I've altered the restriction on nd_assay_genotype
> from UNIQUE(genotype_id) to UNIQUE(genotype_id, nd_assay_id).
>
> I can't see any compatibility issues with the old schema, so does
> anyone have any objections if I roll this into the main source?
>
> -s
>
>
>
>
> --
> Seth Redmond
>   Scientific Programmer, VectorBase
>   Kafatos / Christophides Groups
>   Div. Cell and Molecular Biology
>   Imperial College, London
> [hidden email]
> --
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>



--
Sook Jung, PhD
Assistant Research Professor of Bioinformatics
Dept of Horticulture and Landscape Architecture
Washington State University
45 Johnson Hall, Pullman, WA 99164-6414
Email:[hidden email]

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Pantelis Topalis
In reply to this post by seth redmond
Hi Seth,

I am a bit confused from the paradigm you are using

karyotype is defined as "a standard arrangement of the chromosome complement done for chromosome analysis" and
chromosome complement as "The whole set of chromosomes for the species."

So what do you mean a karyotype for multiple chromosomes ? In humans for example, the karyotype consists of 46 chromosomes and the same is true for the genotype (The genetic constitution (the genome) of a cell, an individual or an organism. )

-P


On 14/7/2010 8:09 μμ, seth redmond wrote:
whilst loading in some test data I encountered a number of cases where two genotypes will arise from a single assay (best example: a single staining will produce a karyotype for multiple chromosomes). As a result, for my DB, I've altered the restriction on nd_assay_genotype from UNIQUE(genotype_id) to UNIQUE(genotype_id, nd_assay_id).

I can't see any compatibility issues with the old schema, so does anyone have any objections if I roll this into the main source?

-s






------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

seth redmond
The karyotypes in this case are the inversions that are present or not in the individual. i.e. a single staining will define a mosquito as 2LA/+ and 2Rbc/bc. hence we need a one-to-many relationship instead of a one-to-one

Kitsos, this is a discussion about the internal workings of the database and as such the semantics are almost entirely unimportant, but if we are to be more precise, a single staining will produce a measurement of a multiple genotypes. 






On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:

Hi Seth,

I am a bit confused from the paradigm you are using

karyotype is defined as "a standard arrangement of the chromosome complement done for chromosome analysis" and
chromosome complement as "The whole set of chromosomes for the species."

So what do you mean a karyotype for multiple chromosomes ? In humans for example, the karyotype consists of 46 chromosomes and the same is true for the genotype (The genetic constitution (the genome) of a cell, an individual or an organism. )

-P


On 14/7/2010 8:09 μμ, seth redmond wrote:
whilst loading in some test data I encountered a number of cases where two genotypes will arise from a single assay (best example: a single staining will produce a karyotype for multiple chromosomes). As a result, for my DB, I've altered the restriction on nd_assay_genotype from UNIQUE(genotype_id) to UNIQUE(genotype_id, nd_assay_id).

I can't see any compatibility issues with the old schema, so does anyone have any objections if I roll this into the main source?

-s







------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Pantelis Topalis
The karyotype refers to the chromosomal arrangement of the whole set of chromosomes for an individual.
You can have an individual 2LA/+ and 2Rbc/bc (and any combinations possible for the third chromosome) and this is a single karyotype (genotype) and not 2 or more karyotypes (genotypes). Therefore in this case a one-to-one relation is sufficient.

-P


On 15/7/2010 12:20 μμ, seth redmond wrote:
The karyotypes in this case are the inversions that are present or not in the individual. i.e. a single staining will define a mosquito as 2LA/+ and 2Rbc/bc. hence we need a one-to-many relationship instead of a one-to-one

Kitsos, this is a discussion about the internal workings of the database and as such the semantics are almost entirely unimportant, but if we are to be more precise, a single staining will produce a measurement of a multiple genotypes. 






On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:

Hi Seth,

I am a bit confused from the paradigm you are using

karyotype is defined as "a standard arrangement of the chromosome complement done for chromosome analysis" and
chromosome complement as "The whole set of chromosomes for the species."

So what do you mean a karyotype for multiple chromosomes ? In humans for example, the karyotype consists of 46 chromosomes and the same is true for the genotype (The genetic constitution (the genome) of a cell, an individual or an organism. )

-P


On 14/7/2010 8:09 ΞΌΞΌ, seth redmond wrote:
whilst loading in some test data I encountered a number of cases where two genotypes will arise from a single assay (best example: a single staining will produce a karyotype for multiple chromosomes). As a result, for my DB, I've altered the restriction on nd_assay_genotype from UNIQUE(genotype_id) to UNIQUE(genotype_id, nd_assay_id).

I can't see any compatibility issues with the old schema, so does anyone have any objections if I roll this into the main source?

-s








__________ Information from ESET NOD32 Antivirus, version of virus signature database 5280 (20100715) __________

The message was checked by ESET NOD32 Antivirus.

http://www.eset.com


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

seth redmond
Yes, in theory, but in practice we will want to break this genotype  
down into it's atomic parts. For example, in the popgen samples we  
applied to the chip we want to compare chromosomal arms individually,  
for instance to pull out all cases of the 2Rcbu karyotype and seek  
polymorphisms that are fixed for this inversion. In this case the 2L  
karyotype is wholly irrelevant and rather than having to write scripts  
that calculate whether "2La/a 2Rb/c" and "2La/+ 2Rb/c" are the same,  
we need the database to be able to apply a simple comparator "2Rbc/bc"  
= "2Rbc/bc".

This is far from the only example. A one-to-one causes untold problems  
with high-throughput assays. Should we consider the 400,000 sites that  
are tested in a single SNP-chip as a single genotype? or instead  
should we have to enter 400,000 records in the assay table, each tied  
to a single SNP-genotype?




On 15 Jul 2010, at 10:58, Pantelis Topalis wrote:

> The karyotype refers to the chromosomal arrangement of the whole set  
> of chromosomes for an individual.
> You can have an individual 2LA/+ and 2Rbc/bc (and any combinations  
> possible for the third chromosome) and this is a single karyotype  
> (genotype) and not 2 or more karyotypes (genotypes). Therefore in  
> this case a one-to-one relation is sufficient.
>
> -P
>
>
> On 15/7/2010 12:20 μμ, seth redmond wrote:
>>
>> The karyotypes in this case are the inversions that are present or  
>> not in the individual. i.e. a single staining will define a  
>> mosquito as 2LA/+ and 2Rbc/bc. hence we need a one-to-many  
>> relationship instead of a one-to-one
>>
>> Kitsos, this is a discussion about the internal workings of the  
>> database and as such the semantics are almost entirely unimportant,  
>> but if we are to be more precise, a single staining will produce a  
>> measurement of a multiple genotypes.
>>
>>
>>
>>
>>
>>
>> On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:
>>
>>> Hi Seth,
>>>
>>> I am a bit confused from the paradigm you are using
>>>
>>> karyotype is defined as "a standard arrangement of the chromosome  
>>> complement done for chromosome analysis" and
>>> chromosome complement as "The whole set of chromosomes for the  
>>> species."
>>>
>>> So what do you mean a karyotype for multiple chromosomes ? In  
>>> humans for example, the karyotype consists of 46 chromosomes and  
>>> the same is true for the genotype (The genetic constitution (the  
>>> genome) of a cell, an individual or an organism. )
>>>
>>> -P
>>>
>>>
>>> On 14/7/2010 8:09 ΞΌΞΌ, seth redmond wrote:
>>>>
>>>> whilst loading in some test data I encountered a number of cases  
>>>> where two genotypes will arise from a single assay (best example:  
>>>> a single staining will produce a karyotype for multiple  
>>>> chromosomes). As a result, for my DB, I've altered the  
>>>> restriction on nd_assay_genotype from UNIQUE(genotype_id) to  
>>>> UNIQUE(genotype_id, nd_assay_id).
>>>>
>>>> I can't see any compatibility issues with the old schema, so does  
>>>> anyone have any objections if I roll this into the main source?
>>>>
>>>> -s
>>>>
>>>>
>>>>
>>>>
>>>
>>
>>
>>
>> __________ Information from ESET NOD32 Antivirus, version of virus  
>> signature database 5280 (20100715) __________
>>
>> The message was checked by ESET NOD32 Antivirus.
>>
>> http://www.eset.com
>


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Pantelis Topalis
  I am still not getting why querying the database for 2Rbc/bc it will
not pick both "2La/a 2Rbc/bc" and "2La/+ 2Rbc/bc" if they are present
and you will need extra scripts for this, but I can see that if you want
to query for combinations of invertions that it may be simpler to have
them stored separately maintaining the information that those are found
on the same individual (genotype). From the genetical point of view this
separation seems a bit awkward so please send us a copy of the database
populated with some data in order to look through them together with
Emmanuel next week.

The SNP set I am carrying is part of my (single) genotype and the same
is true for all the alleles and mutations I am carrying. I don't know
how to represent this in a database but it is impossible to claim that I
have 400,000 SNP-genotypes.

-P

PS. I am getting my emails through gmod-schema list , so I suppose the
list is slow.


On 15/7/2010 1:32 μμ, seth redmond wrote:

> Yes, in theory, but in practice we will want to break this genotype
> down into it's atomic parts. For example, in the popgen samples we
> applied to the chip we want to compare chromosomal arms individually,
> for instance to pull out all cases of the 2Rcbu karyotype and seek
> polymorphisms that are fixed for this inversion. In this case the 2L
> karyotype is wholly irrelevant and rather than having to write scripts
> that calculate whether "2La/a 2Rb/c" and "2La/+ 2Rb/c" are the same,
> we need the database to be able to apply a simple comparator "2Rbc/bc"
> = "2Rbc/bc".
>
> This is far from the only example. A one-to-one causes untold problems
> with high-throughput assays. Should we consider the 400,000 sites that
> are tested in a single SNP-chip as a single genotype? or instead
> should we have to enter 400,000 records in the assay table, each tied
> to a single SNP-genotype?
>
>
>
>
> On 15 Jul 2010, at 10:58, Pantelis Topalis wrote:
>
>> The karyotype refers to the chromosomal arrangement of the whole set
>> of chromosomes for an individual.
>> You can have an individual 2LA/+ and 2Rbc/bc (and any combinations
>> possible for the third chromosome) and this is a single karyotype
>> (genotype) and not 2 or more karyotypes (genotypes). Therefore in
>> this case a one-to-one relation is sufficient.
>>
>> -P
>>
>>
>> On 15/7/2010 12:20 ΞΌΞΌ, seth redmond wrote:
>>>
>>> The karyotypes in this case are the inversions that are present or
>>> not in the individual. i.e. a single staining will define a mosquito
>>> as 2LA/+ and 2Rbc/bc. hence we need a one-to-many relationship
>>> instead of a one-to-one
>>>
>>> Kitsos, this is a discussion about the internal workings of the
>>> database and as such the semantics are almost entirely unimportant,
>>> but if we are to be more precise, a single staining will produce a
>>> measurement of a multiple genotypes.
>>>
>>>
>>>
>>>
>>>
>>>
>>> On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:
>>>
>>>> Hi Seth,
>>>>
>>>> I am a bit confused from the paradigm you are using
>>>>
>>>> karyotype is defined as "a standard arrangement of the chromosome
>>>> complement done for chromosome analysis" and
>>>> chromosome complement as "The whole set of chromosomes for the
>>>> species."
>>>>
>>>> So what do you mean a karyotype for multiple chromosomes ? In
>>>> humans for example, the karyotype consists of 46 chromosomes and
>>>> the same is true for the genotype (The genetic constitution (the
>>>> genome) of a cell, an individual or an organism. )
>>>>
>>>> -P
>>>>
>>>>
>>>> On 14/7/2010 8:09 ����, seth redmond wrote:
>>>>>
>>>>> whilst loading in some test data I encountered a number of cases
>>>>> where two genotypes will arise from a single assay (best example:
>>>>> a single staining will produce a karyotype for multiple
>>>>> chromosomes). As a result, for my DB, I've altered the restriction
>>>>> on nd_assay_genotype from UNIQUE(genotype_id) to
>>>>> UNIQUE(genotype_id, nd_assay_id).
>>>>>
>>>>> I can't see any compatibility issues with the old schema, so does
>>>>> anyone have any objections if I roll this into the main source?
>>>>>
>>>>> -s
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>
>>>
>>>
>>> __________ Information from ESET NOD32 Antivirus, version of virus
>>> signature database 5280 (20100715) __________
>>>
>>> The message was checked by ESET NOD32 Antivirus.
>>>
>>> http://www.eset.com
>>
>
>
> __________ Information from ESET NOD32 Antivirus, version of virus
> signature database 5280 (20100715) __________
>
> The message was checked by ESET NOD32 Antivirus.
>
> http://www.eset.com
>
>
>
>


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

seth redmond
I have already sent you a copy of the database with data in it, the  
sql dump is in my folder on the VBftp site. I'm still waiting for your  
stock tables so I can combine them. The genotypes will quite obviously  
be linked to an individual, since each genotype will be linked, via an  
assay, to the entry in the stock table that defines the individual on  
which it was performed.

If you really object to a one-to-many entry from assay to genotype,  
then please come up with an alternative schema for displaying large  
number of measurements connected to a single assay. I'd suggest doing  
this sooner rather than later, before the DB starts being used in anger.


On 15 Jul 2010, at 12:17, Pantelis Topalis wrote:

>  I am still not getting why querying the database for 2Rbc/bc it will
> not pick both "2La/a 2Rbc/bc" and "2La/+ 2Rbc/bc" if they are present
> and you will need extra scripts for this, but I can see that if you  
> want
> to query for combinations of invertions that it may be simpler to have
> them stored separately maintaining the information that those are  
> found
> on the same individual (genotype). From the genetical point of view  
> this
> separation seems a bit awkward so please send us a copy of the  
> database
> populated with some data in order to look through them together with
> Emmanuel next week.
>
> The SNP set I am carrying is part of my (single) genotype and the same
> is true for all the alleles and mutations I am carrying. I don't know
> how to represent this in a database but it is impossible to claim  
> that I
> have 400,000 SNP-genotypes.
>
> -P
>
> PS. I am getting my emails through gmod-schema list , so I suppose the
> list is slow.
>
>
> On 15/7/2010 1:32 μμ, seth redmond wrote:
>> Yes, in theory, but in practice we will want to break this genotype
>> down into it's atomic parts. For example, in the popgen samples we
>> applied to the chip we want to compare chromosomal arms individually,
>> for instance to pull out all cases of the 2Rcbu karyotype and seek
>> polymorphisms that are fixed for this inversion. In this case the 2L
>> karyotype is wholly irrelevant and rather than having to write  
>> scripts
>> that calculate whether "2La/a 2Rb/c" and "2La/+ 2Rb/c" are the same,
>> we need the database to be able to apply a simple comparator "2Rbc/
>> bc"
>> = "2Rbc/bc".
>>
>> This is far from the only example. A one-to-one causes untold  
>> problems
>> with high-throughput assays. Should we consider the 400,000 sites  
>> that
>> are tested in a single SNP-chip as a single genotype? or instead
>> should we have to enter 400,000 records in the assay table, each tied
>> to a single SNP-genotype?
>>
>>
>>
>>
>> On 15 Jul 2010, at 10:58, Pantelis Topalis wrote:
>>
>>> The karyotype refers to the chromosomal arrangement of the whole set
>>> of chromosomes for an individual.
>>> You can have an individual 2LA/+ and 2Rbc/bc (and any combinations
>>> possible for the third chromosome) and this is a single karyotype
>>> (genotype) and not 2 or more karyotypes (genotypes). Therefore in
>>> this case a one-to-one relation is sufficient.
>>>
>>> -P
>>>
>>>
>>> On 15/7/2010 12:20 ΞΌΞΌ, seth redmond wrote:
>>>>
>>>> The karyotypes in this case are the inversions that are present or
>>>> not in the individual. i.e. a single staining will define a  
>>>> mosquito
>>>> as 2LA/+ and 2Rbc/bc. hence we need a one-to-many relationship
>>>> instead of a one-to-one
>>>>
>>>> Kitsos, this is a discussion about the internal workings of the
>>>> database and as such the semantics are almost entirely unimportant,
>>>> but if we are to be more precise, a single staining will produce a
>>>> measurement of a multiple genotypes.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:
>>>>
>>>>> Hi Seth,
>>>>>
>>>>> I am a bit confused from the paradigm you are using
>>>>>
>>>>> karyotype is defined as "a standard arrangement of the chromosome
>>>>> complement done for chromosome analysis" and
>>>>> chromosome complement as "The whole set of chromosomes for the
>>>>> species."
>>>>>
>>>>> So what do you mean a karyotype for multiple chromosomes ? In
>>>>> humans for example, the karyotype consists of 46 chromosomes and
>>>>> the same is true for the genotype (The genetic constitution (the
>>>>> genome) of a cell, an individual or an organism. )
>>>>>
>>>>> -P
>>>>>
>>>>>
>>>>> On 14/7/2010 8:09 ����, seth redmond wrote:
>>>>>>
>>>>>> whilst loading in some test data I encountered a number of cases
>>>>>> where two genotypes will arise from a single assay (best example:
>>>>>> a single staining will produce a karyotype for multiple
>>>>>> chromosomes). As a result, for my DB, I've altered the  
>>>>>> restriction
>>>>>> on nd_assay_genotype from UNIQUE(genotype_id) to
>>>>>> UNIQUE(genotype_id, nd_assay_id).
>>>>>>
>>>>>> I can't see any compatibility issues with the old schema, so does
>>>>>> anyone have any objections if I roll this into the main source?
>>>>>>
>>>>>> -s
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> __________ Information from ESET NOD32 Antivirus, version of virus
>>>> signature database 5280 (20100715) __________
>>>>
>>>> The message was checked by ESET NOD32 Antivirus.
>>>>
>>>> http://www.eset.com
>>>
>>
>>
>> __________ Information from ESET NOD32 Antivirus, version of virus
>> signature database 5280 (20100715) __________
>>
>> The message was checked by ESET NOD32 Antivirus.
>>
>> http://www.eset.com
>>
>>
>>
>>
>


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: [Popgen] nd_assay_genotype

Pantelis Topalis
  I had the impression that Emmanuel has already sent you a copy of our
database, I will have to check with him. I don't really object to the
change from one to many, I only said that if you are using the karyotype
as one entity a one to one relation is sufficient. If you break it into
specific components then my only requirement is to be able to
reconstruct the karyotype from the atomic components as the last column
of Nora's data does.

-P

On 15/7/2010 2:37 μμ, seth redmond wrote:

> I have already sent you a copy of the database with data in it, the
> sql dump is in my folder on the VBftp site. I'm still waiting for your
> stock tables so I can combine them. The genotypes will quite obviously
> be linked to an individual, since each genotype will be linked, via an
> assay, to the entry in the stock table that defines the individual on
> which it was performed.
>
> If you really object to a one-to-many entry from assay to genotype,
> then please come up with an alternative schema for displaying large
> number of measurements connected to a single assay. I'd suggest doing
> this sooner rather than later, before the DB starts being used in anger.
>
>
> On 15 Jul 2010, at 12:17, Pantelis Topalis wrote:
>
>> I am still not getting why querying the database for 2Rbc/bc it will
>> not pick both "2La/a 2Rbc/bc" and "2La/+ 2Rbc/bc" if they are present
>> and you will need extra scripts for this, but I can see that if you want
>> to query for combinations of invertions that it may be simpler to have
>> them stored separately maintaining the information that those are found
>> on the same individual (genotype). From the genetical point of view this
>> separation seems a bit awkward so please send us a copy of the database
>> populated with some data in order to look through them together with
>> Emmanuel next week.
>>
>> The SNP set I am carrying is part of my (single) genotype and the same
>> is true for all the alleles and mutations I am carrying. I don't know
>> how to represent this in a database but it is impossible to claim that I
>> have 400,000 SNP-genotypes.
>>
>> -P
>>
>> PS. I am getting my emails through gmod-schema list , so I suppose the
>> list is slow.
>>
>>
>> On 15/7/2010 1:32 ΞΌΞΌ, seth redmond wrote:
>>> Yes, in theory, but in practice we will want to break this genotype
>>> down into it's atomic parts. For example, in the popgen samples we
>>> applied to the chip we want to compare chromosomal arms individually,
>>> for instance to pull out all cases of the 2Rcbu karyotype and seek
>>> polymorphisms that are fixed for this inversion. In this case the 2L
>>> karyotype is wholly irrelevant and rather than having to write scripts
>>> that calculate whether "2La/a 2Rb/c" and "2La/+ 2Rb/c" are the same,
>>> we need the database to be able to apply a simple comparator "2Rbc/bc"
>>> = "2Rbc/bc".
>>>
>>> This is far from the only example. A one-to-one causes untold problems
>>> with high-throughput assays. Should we consider the 400,000 sites that
>>> are tested in a single SNP-chip as a single genotype? or instead
>>> should we have to enter 400,000 records in the assay table, each tied
>>> to a single SNP-genotype?
>>>
>>>
>>>
>>>
>>> On 15 Jul 2010, at 10:58, Pantelis Topalis wrote:
>>>
>>>> The karyotype refers to the chromosomal arrangement of the whole set
>>>> of chromosomes for an individual.
>>>> You can have an individual 2LA/+ and 2Rbc/bc (and any combinations
>>>> possible for the third chromosome) and this is a single karyotype
>>>> (genotype) and not 2 or more karyotypes (genotypes). Therefore in
>>>> this case a one-to-one relation is sufficient.
>>>>
>>>> -P
>>>>
>>>>
>>>> On 15/7/2010 12:20 ����, seth redmond wrote:
>>>>>
>>>>> The karyotypes in this case are the inversions that are present or
>>>>> not in the individual. i.e. a single staining will define a mosquito
>>>>> as 2LA/+ and 2Rbc/bc. hence we need a one-to-many relationship
>>>>> instead of a one-to-one
>>>>>
>>>>> Kitsos, this is a discussion about the internal workings of the
>>>>> database and as such the semantics are almost entirely unimportant,
>>>>> but if we are to be more precise, a single staining will produce a
>>>>> measurement of a multiple genotypes.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On 15 Jul 2010, at 08:14, Pantelis Topalis wrote:
>>>>>
>>>>>> Hi Seth,
>>>>>>
>>>>>> I am a bit confused from the paradigm you are using
>>>>>>
>>>>>> karyotype is defined as "a standard arrangement of the chromosome
>>>>>> complement done for chromosome analysis" and
>>>>>> chromosome complement as "The whole set of chromosomes for the
>>>>>> species."
>>>>>>
>>>>>> So what do you mean a karyotype for multiple chromosomes ? In
>>>>>> humans for example, the karyotype consists of 46 chromosomes and
>>>>>> the same is true for the genotype (The genetic constitution (the
>>>>>> genome) of a cell, an individual or an organism. )
>>>>>>
>>>>>> -P
>>>>>>
>>>>>>
>>>>>> On 14/7/2010 8:09 Ξ�οΏ½Ξ�οΏ½Ξ�οΏ½Ξ�οΏ½, seth redmond wrote:
>>>>>>>
>>>>>>> whilst loading in some test data I encountered a number of cases
>>>>>>> where two genotypes will arise from a single assay (best example:
>>>>>>> a single staining will produce a karyotype for multiple
>>>>>>> chromosomes). As a result, for my DB, I've altered the restriction
>>>>>>> on nd_assay_genotype from UNIQUE(genotype_id) to
>>>>>>> UNIQUE(genotype_id, nd_assay_id).
>>>>>>>
>>>>>>> I can't see any compatibility issues with the old schema, so does
>>>>>>> anyone have any objections if I roll this into the main source?
>>>>>>>
>>>>>>> -s
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> __________ Information from ESET NOD32 Antivirus, version of virus
>>>>> signature database 5280 (20100715) __________
>>>>>
>>>>> The message was checked by ESET NOD32 Antivirus.
>>>>>
>>>>> http://www.eset.com
>>>>
>>>
>>>
>>> __________ Information from ESET NOD32 Antivirus, version of virus
>>> signature database 5280 (20100715) __________
>>>
>>> The message was checked by ESET NOD32 Antivirus.
>>>
>>> http://www.eset.com
>>>
>>>
>>>
>>>
>>
>
>
> __________ Information from ESET NOD32 Antivirus, version of virus
> signature database 5281 (20100715) __________
>
> The message was checked by ESET NOD32 Antivirus.
>
> http://www.eset.com
>
>
>
>


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema