Questions about phenotype.assay_id, feature_genotype table and dbxref

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Questions about phenotype.assay_id, feature_genotype table and dbxref


Hello chadoers,

I haven't be able to infer the meaning of some Chado fields from the documentation. This is where I'm struggling with:

1.- What is the phenotype.assay_id for?

2.- In the feature_genotype table, I'm not really sure about what is stored in cvterm_id, rank and cgroup fields.

3.- And a more general question, I'm also interested in understanding how dbxref tables works. I'm wondering if there is extra documentation on this topic apart from the Chado wiki.

As a background of these questions I present here our use case. We will be glad to reading your suggestions or comments.

Use case:

We use the ND module to store phenotypic data from strawberry transgenic lines.

- Stocks:

We define 4 types of stock: cultivar, collection, transgenic line and sample.

  • Cultivars are the strains that were transformed to produce the transgenic lines and therefore, are used as the control group.

  • Collection are all the transgenic lines that were transformed with the same transgene.

  • A transgenic line is plant that came from an independent genetic transformation event. A transgenic line can have multiple plants originated from in vitro culture micro-propagation.

  • In order to store the phenotype values independently, we define the “sample” as each fruit, leaf of plant that was phenotyped. All the samples are related with its corresponding transgenic line through stock_relationship.

- Phenotypes:

Phenotypes are stored in phenotype table and are related to samples by nd_expermient table. A sample gives rise to several phenotypic traits (shape, weight, fruit firmness, color, etc), each of them representing an individual entry in the nd_experiment table. Protocols are related to each phenotyping experiment through nd_experiment_protocol table.

As we have data of transgenic lines from different seasons, we store the year of harvest in nd_experiment_stockprop and the date of the phenotypic assay in nd_experimentprop.

- Genotypes:

We don't have genome data from strawberry but we have some transgene sequences. Maybe in the future we could use a more mature version of the ones currently published. We relate stocks with genotypes using two different strategies.

The fist one involves the nd_experiment table. For each transformation event a nd_experiment entry is created and related to a unique genotype, which is connected to the transgene sequence (feature table) through feature_genotype table. As the sequence is the same for a collection of GMO lines, a unique feature entry is connected to multiple genotypes. We create a new genotype for each line carrying a transgene even though they carry the same sequence, inasmuch as during the transformation process, the construction will be randomly inserted into the genome so we can assume they are not genetically identical. In this case we add 2 entries to the nd_experiment_stock table, one for the parental line (cultivar) with a type_id of 'assay uses stock' and another one for the generated transgenic line with a type_id of 'assay creates stock'.

Sample stocks are also related to genotype and feature tables since they carry the transgene. This connection is made using the stock_genotype table.


Gmod-schema mailing list
[hidden email]