diff options
| author | Pjotr Prins <pjotr.public01@thebird.nl> | 2026-11-15 03:34:34 +0000 |
|---|---|---|
| committer | Pjotr Prins <pjotr.public01@thebird.nl> | 2026-11-15 03:34:34 +0000 |
| commit | 4f2117ce6bbd065059775eaebd98b00135159847 (patch) | |
| tree | 50336d87d91d0c178c14cd3e272d8bf4762844dd /src | |
| parent | c8ce261f2bf499ab3260ff5ce186c801491b9fea (diff) | |
| download | genenetwork4-4f2117ce6bbd065059775eaebd98b00135159847.tar.gz | |
correlation: phenotype traits and datasets, the common case
Either side of a correlation may now be a Publish (phenotype) dataset, not just a ProbeSet one: a pheno trait correlated against an expression dataset is what GN2's mapping page is usually used for, and returning 404 for BXDPublish was the wrong answer. A publish reference trait is addressed by its numeric PublishXRef.Id -- that is how GN2 addresses phenotype traits, and a name there is refused with a message saying so rather than a confusing "no such trait". The bulk fetch gains the publish equivalent, keyed on the freeze's InbredSetId, so the same individuals line up on both sides. Measured live: BXDPublish/10001 (34 observations) against HC_M2_0606_P, top 5, HTTP 200 in 5.0 s; 1440079_at at r -0.757 (p 3.2e-6) leads.
Diffstat (limited to 'src')
| -rw-r--r-- | src/db.rs | 37 | ||||
| -rw-r--r-- | src/routes/correlation.rs | 87 |
2 files changed, 101 insertions, 23 deletions
diff --git a/src/db.rs b/src/db.rs index 7464125..91734f7 100644 --- a/src/db.rs +++ b/src/db.rs @@ -1594,6 +1594,43 @@ pub fn get_dataset_values_for_individuals( .map_err(DbError::from) } +/// The same bulk fetch for a Publish (phenotype) dataset. +/// +/// `PublishXRef.Id` is the trait's handle within that dataset -- GN2 addresses +/// phenotype traits by it ("1427571_at" style names belong to the expression +/// sets) -- and the values live in `PublishData`, keyed by the same InbredSetId +/// as the reference trait's group. Returned as (trait id, individual, value) +/// so the caller can group it exactly like the probeset case. +pub fn get_publish_values_for_individuals( + inbred_set_id: u16, + individuals: &[String], +) -> Result<Vec<(String, String, Option<f64>)>, DbError> { + if individuals.is_empty() { + return Ok(Vec::new()); + } + let pool = pool()?; + let mut conn = pool.get_conn()?; + let placeholders = std::iter::repeat_n("?", individuals.len()) + .collect::<Vec<_>>() + .join(","); + let query = format!( + "SELECT CAST(PublishXRef.Id AS CHAR), Strain.Name, PublishData.value \ + FROM PublishData, Strain, PublishXRef \ + WHERE PublishXRef.InbredSetId = ? AND PublishXRef.DataId = PublishData.Id \ + AND PublishData.StrainId = Strain.Id \ + AND Strain.Name IN ({placeholders}) \ + ORDER BY PublishXRef.Id, Strain.Name" + ); + let mut params = mysql::Params::Positional(vec![mysql::Value::from(inbred_set_id)]); + if let mysql::Params::Positional(values) = &mut params { + values.extend(individuals.iter().map(|name| mysql::Value::from(name.as_str()))); + } + conn.exec_map(&query, params, |(trait_id, individual, value)| { + (trait_id, individual, value) + }) + .map_err(DbError::from) +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/routes/correlation.rs b/src/routes/correlation.rs index 36faedb..00442a9 100644 --- a/src/routes/correlation.rs +++ b/src/routes/correlation.rs @@ -163,23 +163,17 @@ impl Failure { } /// The blocking half: resolve both datasets, fetch the values, correlate. +/// +/// Either side may be a ProbeSet (expression) dataset or a Publish (phenotype) +/// one: GN2's `corr_dataset` menu mixes them, so a pheno trait correlated +/// against an expression dataset -- the common case -- has to work. fn compute( request: &CorrelationRequest, method: Method, top_n: usize, ) -> Result<(String, Vec<correlation::Hit>), Failure> { - let reference_dataset = db::get_dataset_by_name(&request.dataset) - .map_err(|error| Failure::Database(error.to_string()))? - .ok_or_else(|| Failure::NotFound(format!("no dataset named '{}'", request.dataset)))?; - - let values = db::get_dataset_probeset_values(reference_dataset.id, &request.trait_name) - .map_err(|error| Failure::Database(error.to_string()))?; - let reference = correlation::Reference::from_pairs( - values - .into_iter() - .map(|value| (value.name, value.value)), - ); - if reference.is_empty() { + let reference = resolve_reference(request)?; + if reference.values.is_empty() { return Err(Failure::NoValues(format!( "'{}' has no values in '{}', so there is nothing to correlate", request.trait_name, request.dataset @@ -194,17 +188,7 @@ fn compute( ))); } - let target = db::get_dataset_by_name(&request.target_dataset) - .map_err(|error| Failure::Database(error.to_string()))? - .ok_or_else(|| { - Failure::NotFound(format!( - "no dataset named '{}'", - request.target_dataset - )) - })?; - - let rows = db::get_dataset_values_for_individuals(target.id, &reference.names) - .map_err(|error| Failure::Database(error.to_string()))?; + let rows = resolve_target_rows(&request.target_dataset, &reference.names)?; let hits = correlation::correlate_many(&reference, method, rows, top_n); Ok(( format!( @@ -216,3 +200,60 @@ fn compute( hits, )) } + +/// The reference trait's values, from whichever kind of dataset holds it. +fn resolve_reference(request: &CorrelationRequest) -> Result<correlation::Reference, Failure> { + if let Some(dataset) = + db::get_dataset_by_name(&request.dataset).map_err(db_failure)? + { + let values = db::get_dataset_probeset_values(dataset.id, &request.trait_name) + .map_err(db_failure)?; + return Ok(correlation::Reference::from_pairs( + values.into_iter().map(|value| (value.name, value.value)), + )); + } + if let Some(publish) = + db::get_publish_freeze_by_name(&request.dataset).map_err(db_failure)? + { + // A publish trait is addressed by its numeric id within the set. + let trait_id: u32 = request.trait_name.trim().parse().map_err(|_| { + Failure::NoValues(format!( + "'{}' is a phenotype dataset, so its traits are addressed by id (a number), \ + not by name '{}'", + request.dataset, request.trait_name + )) + })?; + let values = db::get_group_trait_values(publish.inbred_set_id, trait_id) + .map_err(db_failure)?; + let values: Vec<(String, Option<f64>)> = values + .into_iter() + .map(|value| (value.name, value.value)) + .collect(); + return Ok(correlation::Reference::from_pairs(values)); + } + Err(Failure::NotFound(format!( + "no dataset named '{}'", + request.dataset + ))) +} + +/// Every candidate trait's values in the target dataset, restricted to the +/// reference's individuals. +fn resolve_target_rows( + dataset: &str, + individuals: &[String], +) -> Result<Vec<(String, String, Option<f64>)>, Failure> { + if let Some(target) = db::get_dataset_by_name(dataset).map_err(db_failure)? { + return db::get_dataset_values_for_individuals(target.id, individuals) + .map_err(db_failure); + } + if let Some(publish) = db::get_publish_freeze_by_name(dataset).map_err(db_failure)? { + return db::get_publish_values_for_individuals(publish.inbred_set_id, individuals) + .map_err(db_failure); + } + Err(Failure::NotFound(format!("no dataset named '{dataset}'"))) +} + +fn db_failure(error: db::DbError) -> Failure { + Failure::Database(error.to_string()) +} |
