eaf method and visualizer

Author: colinvwood

q2_qsip2/plugin_setup.py

@@ -8,17 +8,18 @@
 
 import importlib
 
-from qiime2.plugin import Citations, Int, List, Metadata, Plugin, Str
+from qiime2.plugin import Citations, Float, Int, List, Metadata, Plugin, Str
 from q2_types.feature_table import FeatureTable, Frequency
 
 from q2_qsip2 import __version__
 from q2_qsip2.types import QSIP2Data, Unfiltered, Filtered, EAF
 from q2_qsip2.workflow import (
-    standard_workflow, create_qsip_data, subset_and_filter
+    standard_workflow, create_qsip_data, subset_and_filter,
+    resample_and_calculate_EAF
 )
 from q2_qsip2.visualizers._visualizers import (
     plot_weighted_average_densities, plot_sample_curves, plot_density_outliers,
-    show_comparison_groups, plot_filtered_features
+    show_comparison_groups, plot_filtered_features, plot_excess_atom_fractions
 )
 
 
@@ -34,34 +35,9 @@
         "data."
     ),
     short_description="Analyze qSIP data.",
-    # TODO
     citations=[citations['Caporaso-Bolyen-2024']]
 )
 
-plugin.methods.register_function(
-    function=standard_workflow,
-    inputs={
-        'table': FeatureTable[Frequency],
-        'qsip_metadata': QSIP2Data[Unfiltered],
-    },
-    parameters={},
-    outputs=[
-        ('output_table', FeatureTable[Frequency])
-    ],
-    input_descriptions={
-        'table': 'The feature table.',
-        'qsip_metadata': 'The qSIP metadata.',
-    },
-    parameter_descriptions={},
-    output_descriptions={
-        'output_table': 'Placeholder.'
-    },
-    name='Run the standard qSIP2 workflow.',
-    description=(
-        'Placeholder.'
-    )
-)
-
 plugin.methods.register_function(
     function=create_qsip_data,
     inputs={
@@ -99,7 +75,8 @@
     name='Bundle your qSIP metadata and feature table.',
     description=(
         'Placeholder.'
-    )
+    ),
+    citations=[]
 )
 
 plugin.methods.register_function(
@@ -147,7 +124,40 @@
     name='Subset sources and filter features to prepare for comparison.',
     description=(
         'Placeholder.'
-    )
+    ),
+    citations=[]
+)
+
+plugin.methods.register_function(
+    function=resample_and_calculate_EAF,
+    inputs={
+        'filtered_qsip_data': QSIP2Data[Filtered]
+    },
+    parameters={
+        'resamples': Int,
+        'random_seed': Int,
+    },
+    outputs=[
+        ('eaf_qsip_data', QSIP2Data[EAF])
+    ],
+    input_descriptions={
+        'filtered_qsip_data': 'Your filtered qSIP2 data.'
+    },
+    parameter_descriptions={
+        'resamples': 'The number of bootstrap resamplings to perform.',
+        'random_seed': 'The random seed to use during resampling.',
+    },
+    output_descriptions={
+        'eaf_qsip_data': (
+            'Your qSIP2 data with excess atom fraction (EAF) values '
+            'calculated on a per-taxon basis.'
+        )
+    },
+    name='Calculate excess atom fraction (EAF).',
+    description=(
+        'Placeholder.'
+    ),
+    citations=[]
 )
 
 plugin.visualizers.register_function(
@@ -249,4 +259,34 @@
     citations=[],
 )
 
+plugin.visualizers.register_function(
+    function=plot_excess_atom_fractions,
+    inputs={
+        'eaf_qsip_data': QSIP2Data[EAF],
+    },
+    parameters={
+        'num_top': Int,
+        'confidence_interval': Float
+    },
+    input_descriptions={
+        'eaf_qsip_data': 'Your EAF-calculated qSIP2 data.',
+    },
+    parameter_descriptions={
+        'num_top': (
+            'The number of taxa displayed, selected in order of decreasing '
+            'excess atom fraction.'
+        ),
+        'confidence_interval': (
+            'The confidence interval to display from the bootstrapped excess '
+            'atom fractions.'
+        )
+    },
+    name='Visualize per-taxon excess atom fractions.',
+    description=(
+        'Plots per-taxon excess atom fractions with bootstrapped confidence '
+        'intervals.'
+    ),
+    citations=[]
+)
+
 importlib.import_module('q2_qsip2.types._deferred_setup')

q2_qsip2/types/_deferred_setup/_transformers.py

@@ -54,7 +54,7 @@ def _4(ff: QSIP2DataFilteredFormat) -> RS4:
 
 @plugin.register_transformer
 def _5(qsip_object: RS4) -> QSIP2DataEAFFormat:
-    ff = QSIP2DataFilteredFormat()
+    ff = QSIP2DataEAFFormat()
     return _qsip_object_to_format(qsip_object, ff)
 
 

q2_qsip2/visualizers/_visualizers.py

@@ -124,3 +124,34 @@ def plot_filtered_features(output_dir: str, filtered_qsip_data: RS4) -> None:
     _ggplot2_object_to_visualization(
         plot, Path(output_dir), width=10, height=10
     )
+
+
+def plot_excess_atom_fractions(
+    output_dir: str,
+    eaf_qsip_data: RS4,
+    num_top: int,
+    confidence_interval: float = 0.9
+) -> None:
+    '''
+    Plots per-taxon excess atom fraction values.
+
+    Parameters
+    ----------
+    output_dir : str
+        The root directory of the visualization loaded into the browser.
+    qsip_data : RS4
+        The "qsip_data" object.
+    num_top : int
+        The number of taxa displayed taken in order of decreasing excess
+        atom fraction.
+    confidence_interval : float
+        The confidence interval to display from the bootstrapped excess atom
+        fraction values.
+    '''
+    plot = qsip2.plot_EAF_values(
+        eaf_qsip_data, top=num_top, confidence=confidence_interval, error='bar'
+    )
+
+    _ggplot2_object_to_visualization(
+        plot, Path(output_dir), width=10, height=10
+    )

q2_qsip2/workflow.py

@@ -127,10 +127,10 @@ def subset_and_filter(
     qsip_data: RS4,
     unlabeled_sources: list[str],
     labeled_sources: list[str],
-    min_unlabeled_sources: int,
-    min_labeled_sources: int,
-    min_unlabeled_fractions: int,
-    min_labeled_fractions: int
+    min_unlabeled_sources: int = 1,
+    min_labeled_sources: int = 1,
+    min_unlabeled_fractions: int = 1,
+    min_labeled_fractions: int = 1
 ) -> RS4:
     '''
     Subsets the qsip data object to include only those sources listed in
@@ -158,14 +158,45 @@ def subset_and_filter(
         The minimum number of fractions a feature must be present in
         to be considered present in a labeled source.
     '''
+    unlabeled_sources_vector = ro.vectors.StrVector(unlabeled_sources)
+    labeled_sources_vector = ro.vectors.StrVector(labeled_sources)
+
     filtered_qsip_data = qsip2.run_feature_filter(
         qsip_data,
-        unlabeled_source_mat_ids=unlabeled_sources,
-        labeled_source_mat_ids=labeled_sources,
+        unlabeled_source_mat_ids=unlabeled_sources_vector,
+        labeled_source_mat_ids=labeled_sources_vector,
         min_unlabeled_sources=min_unlabeled_sources,
         min_labeled_sources=min_labeled_sources,
         min_unlabeled_fractions=min_unlabeled_fractions,
         min_labeled_fractions=min_labeled_fractions
     )
 
     return filtered_qsip_data
+
+
+def resample_and_calculate_EAF(
+    filtered_qsip_data: RS4,
+    resamples: int = 1000,
+    random_seed: int = 1,
+) -> RS4:
+    '''
+    Reseample and calculate excess atom fraction (EAF) for each feature.
+
+    Parameters
+    ----------
+    filtered_qsip_data : RS4
+        The filtered "qsip_data" object.
+    resamples : int
+        The number of bootstrap resamplings to perform.
+    random_seed : int
+        The random seed to use during resampling. Exposed for reproducibility.
+    '''
+    resampled_qsip_data = qsip2.run_resampling(
+        filtered_qsip_data,
+        resamples=resamples,
+        with_seed=random_seed
+    )
+
+    eaf_qsip_data = qsip2.run_EAF_calculations(resampled_qsip_data)
+
+    return eaf_qsip_data