diff options
Diffstat (limited to 'gn3')
-rw-r--r-- | gn3/db/species.py | 31 | ||||
-rw-r--r-- | gn3/partial_correlations.py | 38 |
2 files changed, 68 insertions, 1 deletions
diff --git a/gn3/db/species.py b/gn3/db/species.py index 0deae4e..1e5015f 100644 --- a/gn3/db/species.py +++ b/gn3/db/species.py @@ -30,3 +30,34 @@ def get_chromosome(name: str, is_species: bool, conn: Any) -> Optional[Tuple]: with conn.cursor() as cursor: cursor.execute(_sql) return cursor.fetchall() + +def translate_to_mouse_gene_id(species: str, geneid: int, conn: Any) -> int: + """ + Translate rat or human geneid to mouse geneid + + This is a migration of the + `web.webqtl.correlation/CorrelationPage.translateToMouseGeneID` function in + GN1 + """ + assert species in ("rat", "mouse", "human"), "Invalid species" + if geneid is None: + return 0 + + if species == "mouse": + return geneid + + with conn.cursor as cursor: + if species == "rat": + cursor.execute( + "SELECT mouse FROM GeneIDXRef WHERE rat = %s", geneid) + rat_geneid = cursor.fetchone() + if rat_geneid: + return rat_geneid[0] + + cursor.execute( + "SELECT mouse FROM GeneIDXRef WHERE human = %s", geneid) + human_geneid = cursor.fetchone() + if human_geneid: + return human_geneid[0] + + return 0 # default if all else fails diff --git a/gn3/partial_correlations.py b/gn3/partial_correlations.py index c556d10..1fb0ccc 100644 --- a/gn3/partial_correlations.py +++ b/gn3/partial_correlations.py @@ -6,7 +6,7 @@ GeneNetwork1. """ from functools import reduce -from typing import Any, Sequence +from typing import Any, Tuple, Sequence def control_samples(controls: Sequence[dict], sampleslist: Sequence[str]): """ @@ -86,3 +86,39 @@ def fix_samples(primary_trait: dict, control_traits: Sequence[dict]) -> Sequence control_vals_vars[0], tuple(primary_trait[sample]["variance"] for sample in primary_samples), control_vals_vars[1]) + +def find_identical_traits( + primary_name: str, primary_value: float, control_names: Tuple[str, ...], + control_values: Tuple[float, ...]) -> Tuple[str, ...]: + """ + Find traits that have the same value when the values are considered to + 3 decimal places. + + This is a migration of the + `web.webqtl.correlation.correlationFunction.findIdenticalTraits` function in + GN1. + """ + def __merge_identicals__( + acc: Tuple[str, ...], + ident: Tuple[str, Tuple[str, ...]]) -> Tuple[str, ...]: + return acc + ident[1] + + def __dictify_controls__(acc, control_item): + ckey = "{:.3f}".format(control_item[0]) + return {**acc, ckey: acc.get(ckey, tuple()) + (control_item[1],)} + + return (reduce(## for identical control traits + __merge_identicals__, + (item for item in reduce(# type: ignore[var-annotated] + __dictify_controls__, zip(control_values, control_names), + {}).items() if len(item[1]) > 1), + tuple()) + or + reduce(## If no identical control traits, try primary and controls + __merge_identicals__, + (item for item in reduce(# type: ignore[var-annotated] + __dictify_controls__, + zip((primary_value,) + control_values, + (primary_name,) + control_names), {}).items() + if len(item[1]) > 1), + tuple())) |