Introduction to Chado
What's a chado?
- At first, a database for FlyBase: incredibly interesting dataset
- A database for very deep curation
- An integrated database
- A database that is generic enough to use for any organism
Chado - the way of tea
Why the reference to tea? According to ancient GMOD lore, legend has it that creators Chris Mungall and Dave Emmert were drinking tea in a tea house when they developed the first design that eventually became Chado.
What Documentation Exists for chado?
This Wiki is currently the best source of documentation for Chado. Here are more useful pages in the Wiki:
A Modular Schema
Chado Module List
- Audit - for database audits
- Companalysis - for data from computational analysis
- Contact - for people, groups, and organizations
- Controlled Vocabulary (cv) - for controlled vocabularies and ontologies
- Expression - for summaries of RNA and protein expresssion
- General - for identifiers
- Genetic - for genetic data and genotypes
- Library - for descriptions of molecular libraries
- Mage - for microarray data
- Map - for maps without sequence
- Organism - for taxonomic data
- Phenotype - for phenotypic data
- Phylogeny - for organisms and phylogenetic trees
- Publication (pub) - for publications and references
- Sequence - for sequences and sequence features
- Stock - for specimens and biological collections
- WWW -
There are dependencies between the modules.
The actual chado tables themselves are not discussed in attribute-by-attribute detail; this can be browsed by checking out the Chado Manual.
One of the main strengths of chado is that it brings the sequence and genetics views of the world together.
Let's look at the chado conceptualisation of the world before diving in to the schema design.
The Sequence Module and Features
- A feature is a thingy
- A feature is a potentially localisable
- A feature is further defined by an ontology
Feature Types: an Ontology
One way of representing ontologies is through a graph model, with nodes representing concepts and edges representing relationship types between the concepts.
is_a: subtypes, specialisation/generalization
Some other feature types
- protein domain
- chromosome arm
- regulatory region
- variation features: insertions, deletions, SNPs
The compositional relationships are a subset of the relationships from gene to CDS in SO. How does one decide where to simplify the graph?
The nodes in the graph represent instances of features - the arcs in these cases represent compositional relationships (although other relationship types are possible). Feature graphs do not represent positional or spatial relationships - we will get to that later. For more on representing genes also see Chado Best Practices.
Other cases of trans-splicing may involve spatially distributed primary transcripts.
Feature Graph Transformations
Representing Graphs in a Relational Database
A graph can be defined as a collection of Edges (arcs) and Vertices (nodes).
Representing Ontology Graphs in Chado
The relationship type is a controlled term in itself. Each cvrelationship can be thought of as a SUBJECT PREDICATE OBJECT statement (eg "GPCR is-a transmembrane_receptor).
The structure above is exactly the same as the RDF datamodel - many modern ontology languages (eg DAML, OWL) are layered on top of RDF, so the above structure ensures we will be able to represent all the most advanced ontological concepts.
Representing Feature Graphs in Chado
features are the nodes - feature_relationships are the arcs
Note: the different classes of features could be modeled relationally; the principle is to keep the stable stuff modeled relationally, and the fluid/extensible stuff modeled in an ontology that sits in a generic database structure.
Features are typed
Most implementations of SQL are non-recursive.
Problem: find all genes; find all genes (generic) find all noncoding genes find all protein coding genes find all tRNA genes find all snRNA genes find all snoRNA genes ...etc eek!
Solution: pre-compute transitive closure
GO Ontology subgraph
forall x ALWAYS TRUE: x R* x
x R y IMPLIES: x R* y
x R y, y R* z IMPLIES: x R* z
Using views to simplify queries
The following view uses the cvpath table, which includes the closure of the is_a relationship.
CREATE VIEW fgene AS SELECT feature.* FROM feature INNER JOIN cvpath ON (feature.ftype_id = cvpath.subjterm_id) INNER JOIN cvterm ON (cvpath.objterm_id = cvterm.cvterm_id) WHERE cvterm.termname = 'gene';
At Flybase we will mostly be using chado in data-mining mode - i.e. we will be querying, not updating. This means we can materialize views for speed.
The chado relational model defines a fixed set of attributes for a feature:
|dbxref_id||no||namespaced identifier (foreign key)|
|name||no||A not-guaranteed-unique identifier that is useful to a human|
|uniquename||yes||name or identifier that is unique in the database|
|residues||no||DNA, RNA or protein sequence|
|md5checksum||no||signature of sequence|
|seqlen||no||length of sequence (may be present even if residues is absent)|
|type_id||yes||Sequence Ontology feature type (foreign key)|
What happens if we want to include other attributes specific to certain projects, or specific to certain feature types? We can use an extensible feature property paradigm:
Localising Features in Sequence Coordinates
All sequence localisations are with respect to another feature.
A feature can have multiple locations - however, "split" locations are not used (for an example of a split location, look at how Genbank represents a transcript).
The featureloc table includes the following attributes:
- fmin - 5' boundary of features
- fmax - 3' boundary of features
- strand: the direction of the feature, relative to srcfeature
locations have directional semantics (like mathematical vectors). this is different from the min/max semantics used by e.g. Bioperl, genbank
Chado uses interbase coordinates.
Interbase counts spaces, not bases; starts at zero. More mathematically appealing. Standard (eg genbank, blast) coordinate systems do not allow for proper representation of zero-length features. An example of a zero-length feature is a cleavage site that is between, but does not include, 2 amino acids.
Interbase coordinates (top) and base-oriented (below)
The position of the ATG in interbase is [3, 6] (between the 3rd and 6th gaps)
The position of the ATG in base coordinates is [4, 6] (between 4th and 6th bases inclusive)
Note the different arithmetic for calculating length in these two systems.
Unlike mathematical vectors, we must also explicitly store the directionality (strand). Even though this is surplus to requirements most of the time, it is required for zero-length features, and for circular chromosomes.
Basic example - with locations
Central dogma - with exons and CDSs localised
Using the principle of minimal storage (do not store anything that does not increase the information content of the database - i.e. nothing redundant), we store only exon and CDS boundary localisations. In the Berkeley Drosophila Genome Project (BDGP) data warehouse instantiation of Chado, we may choose to store locations for all features where known - this can vastly simplify some queries, but care must be taken to make sure we don't end up with inconsistent data.
For the most part, infering the boundaries of composite features requires fairly simple graph transformations, although care must be taken for the genes that break central dogma rules.
Locations can be nested
Note that the position of the repeat on the chromosome arm is implicit, and can be calculated with a simple graph transform, but following the principle of minimal storage, we do not store this in the management database.
If we wish to store the redundant position in a for-querying copy of the db, chado allows us this option - we can have as many locations as we like for a feature. We use an extra attribute called locgroup to distinguish locations. locgroup=0 is conventionally used for the non-redundant location.
Computational analysis: Predictions
All predictions are handled analagously to standard central dogma cases. Compute results have scores, and are attached to tuples in the analysis table - this is what distinguishes them from "annotations".
Genscan predicts CDSs and CDS exons (not genes in the Sequence Ontology sense). A typical Genscan prediction may look like this:
Computational analysis: Similarity results
Pairwise similarities (e.g. BLAST) and multiple alignments (e.g. CLUSTAL) are represented differently.
Pairwise alignments produce HSPs. HSPs are scored features with two locations - one on the query, one on the subject.
Computational analysis: Multiple alignments
These are treated analagously to pairwise alignments - just add more locations.
Variation features (e.g. SNPs, insertions, deletions) are treated in a similar fashion to pairwise alignments.
Bioperl and chado mapping
There are important differences between the Bioperl and chado models.
Bioperl allows a feature to have multiple non-contiguous locations. Even though the chado schema allows multiple locations, attaching multiple non-contiguous locations is a violation of the chado semantics. To cope with this, we create extra features for the sublocations. For instance, for a transcript, we would create an exon for each of the sublocations.
If you parse a genbank file including variations (e.g. SNPs) into Bioperl objects, you will get a feature with two properties of type "allele". This can be represented using chado; however, the chado semantics state that these variations should be represented using multiple locations.
- Dave Emmert
- Colin Wiel
- Stan Letovsky
- ShengQuiang Shu
- Pinglei Zhou
- Suzanna Lewis
Chado beta testers and other feedback
- Mark Yandell
- Aubrey de Grey
About this Page
- This page is a Wikified version of a presentation by Chris Mungall.