comparison anndata_operations.xml @ 28:a0274bc43b7e draft

planemo upload for repository https://github.com/ebi-gene-expression-group/container-galaxy-sc-tertiary/tree/develop/tools/tertiary-analysis/scanpy commit d01fa18235ac692874cd3cfddef33696c2df8ac1-dirty

27:7ebc22f77d86

<?xml version="1.0" encoding="utf-8"?>
<tool id="anndata_ops" name="AnnData Operations" version="@TOOL_VERSION@+galaxy91" profile="@PROFILE@">
  <description>modifies metadata and flags genes</description>
  <macros>
    <import>scanpy_macros2.xml</import>
  </macros>
  <expand macro="requirements"/>

python $operations
]]></command>
  <configfiles>
    <configfile name="operations">
import gc
import scanpy as sc
import anndata
from numpy import all
import logging

  if new_field is None:
    new_field = f"{field}_u"
  appendents = (suffix + df.groupby(field).cumcount().astype(str).replace('0','')).replace(suffix, '')
  df[new_field] = df[field].astype(str) + appendents.astype(str)
  return df
	    
adata = sc.read('input.h5')

#if $add_cell_metadata.default:
import pandas as pd

def add_cell_metadata(ad, metadata_file="cell_metadata.tsv", drop_duplicates=True):
  metadata_df = pd.read_csv(metadata_file, sep="\t", index_col=0)

              if prev_dtype == str or prev_dtype == object:
                  prev_dtype = "category"
              print(f"Changing {col} from {merged_obs[col].dtype} to {prev_dtype}")
              merged_obs[col] = merged_obs[col].astype(prev_dtype)
  return merged_obs
	    
adata.obs = add_cell_metadata(adata)
#end if

#if $copy_adata_to_raw:
adata.raw = adata
#end if

gene_name = '${gene_symbols_field}'
qc_vars = list()

if 'n_cells' not in adata.var.columns:
    sc.pp.filter_genes(adata, min_cells=0)
if 'n_counts' not in adata.var.columns:
    sc.pp.filter_genes(adata, min_counts=0)

adata.write('output.h5', compression='gzip')
    </configfile>
</configfiles>

  <inputs>
    <param name="input_obj_file" argument="input-object-file" type="data" format="h5,h5ad" label="Input object in hdf5 AnnData format"/>
    <expand macro="output_object_params_no_loom"/>
    <conditional name="add_cell_metadata">
      <param name="default" type="boolean" checked="false" label="Merge additional cell metadata"/>
      <when value="true">
        <param name="file" type="data" label="Cell metadata with headers" help="A tabular file with headers, where the first column contains cell barcodes. Will be merged via a left join, so not all cells in the obs need to be in the metadata. Currently duplicated column headers will be ignored and the originals in the AnnData will be kept." format="tsv,tabular"/>
      </when>

        </repeat>
        <param name="uns_sources" type="data" label="AnnData objects with uns to copy" help="Extracts uns (such as ranked_genes_groups) from these AnnData objects and merges them into the main input. Make sure to use AnnData objects that are compatible in terms of genes, cells and expression data." format="h5,h5ad" multiple="true"/>
      </when>
    </conditional>
    <param name="sanitize_varm" type="boolean" checked="false" label="Sanitise any null raw.varm objects if any" help="This might be relevant for interfacing with newer versions of AnnData, that might complain if .raw includes a varm null object."/>
  </inputs>

  <outputs>
    <expand macro="output_data_obj_no_loom" description="metadata changes on"/>
  </outputs>

  <tests>
    <test>
      <param name="input_obj_file" value="find_cluster.h5"/>
      <output name="output_h5ad" file="anndata_ops.h5" ftype="h5ad" compare="sim_size"/>
    </test>
    <test>
      <param name="input_obj_file" value="anndata_ops.h5"/>
      <param name="from_var" value = "gene_symbols" />
      <param name="to_var" value = "hello_all" />

      <output name="output_h5ad" file="anndata_ops_layer.h5" ftype="h5ad" compare="sim_size">
        <assert_contents>
          <has_h5_keys keys="layers/filtered" />
        </assert_contents>
      </output>
    </test>
  </tests>

  <help><![CDATA[
=============================

28:a0274bc43b7e

<?xml version="1.0" encoding="utf-8"?>
<tool id="anndata_ops" name="AnnData Operations" version="@TOOL_VERSION@+galaxy92" profile="@PROFILE@">
  <description>modifies metadata and flags genes</description>
  <macros>
    <import>scanpy_macros2.xml</import>
  </macros>
  <expand macro="requirements"/>

python $operations
]]></command>
  <configfiles>
    <configfile name="operations">
import gc
from os import makedirs
import scanpy as sc
import anndata
from numpy import all
import logging

  if new_field is None:
    new_field = f"{field}_u"
  appendents = (suffix + df.groupby(field).cumcount().astype(str).replace('0','')).replace(suffix, '')
  df[new_field] = df[field].astype(str) + appendents.astype(str)
  return df

adata = sc.read('input.h5')
	    
#if $add_cell_metadata.default:
import pandas as pd

def add_cell_metadata(ad, metadata_file="cell_metadata.tsv", drop_duplicates=True):
  metadata_df = pd.read_csv(metadata_file, sep="\t", index_col=0)

              if prev_dtype == str or prev_dtype == object:
                  prev_dtype = "category"
              print(f"Changing {col} from {merged_obs[col].dtype} to {prev_dtype}")
              merged_obs[col] = merged_obs[col].astype(prev_dtype)
  return merged_obs

adata.obs = add_cell_metadata(adata)
#end if

#if $copy_adata_to_raw:
adata.raw = adata
#end if

#if $swap_layer_to_x.default:
#if $swap_layer_to_x.new_name_x:
adata.layers['${swap_layer_to_x.new_name_x}'] = adata.X
#end if
adata.X = adata.layers['${swap_layer_to_x.layer}']
del adata.layers['${swap_layer_to_x.layer}']
#end if

gene_name = '${gene_symbols_field}'
qc_vars = list()

if 'n_cells' not in adata.var.columns:
    sc.pp.filter_genes(adata, min_cells=0)
if 'n_counts' not in adata.var.columns:
    sc.pp.filter_genes(adata, min_counts=0)

#if not $split_on_obs.default or $split_on_obs.output_main:
adata.write('output.h5', compression='gzip')
#end if

#if $split_on_obs.default:
s = 0
res_dir = "output_split"
makedirs(res_dir, exist_ok=True)
for field_value in adata.obs["${split_on_obs.key}"].unique():
    ad_s = adata[adata.obs.${split_on_obs.key} == field_value]
    ad_s.write(f"{res_dir}/${split_on_obs.key}_{s}.h5", compression='gzip')
    if s > 0:
        gc.collect()
    s += 1
#end if


    </configfile>
</configfiles>

  <inputs>
    <param name="input_obj_file" argument="input-object-file" type="data" format="h5,h5ad" label="Input object in hdf5 AnnData format"/>
    <expand macro="output_object_params_no_loom"/>
    <conditional name="swap_layer_to_x">
      <param name="default" type="boolean" checked="false" label="Swap layer to X"/>
      <when value="true">
        <param name="layer" type="text" value="" label="Name of layer to swap to X" help="This layer name needs to exist within ad.layers or this will fail."/>
        <param name="new_name_x" type="text" value="old_X" label="Name of the new slot for X within layers" help="Leave empty and the old X will be lost."/>
      </when>
      <when value="false"/>
    </conditional>
    <conditional name="add_cell_metadata">
      <param name="default" type="boolean" checked="false" label="Merge additional cell metadata"/>
      <when value="true">
        <param name="file" type="data" label="Cell metadata with headers" help="A tabular file with headers, where the first column contains cell barcodes. Will be merged via a left join, so not all cells in the obs need to be in the metadata. Currently duplicated column headers will be ignored and the originals in the AnnData will be kept." format="tsv,tabular"/>
      </when>

        </repeat>
        <param name="uns_sources" type="data" label="AnnData objects with uns to copy" help="Extracts uns (such as ranked_genes_groups) from these AnnData objects and merges them into the main input. Make sure to use AnnData objects that are compatible in terms of genes, cells and expression data." format="h5,h5ad" multiple="true"/>
      </when>
    </conditional>
    <param name="sanitize_varm" type="boolean" checked="false" label="Sanitise any null raw.varm objects if any" help="This might be relevant for interfacing with newer versions of AnnData, that might complain if .raw includes a varm null object."/>
    <conditional name="split_on_obs">
      <param name="default" type="boolean" checked="false" label="Split on obs" help="Split the AnnData object into multiple AnnData objects based on the values of a given obs key. This is useful for example to split a dataset based on a cluster annotation."/>
      <when value="true">
        <param name="key" type="text" label="Obs key to split on" help="The obs key to split on. For example, if you want to split on cluster annotation, you can use the key 'louvain'."/>
        <param name="output_main" type="boolean" checked="true" label="Output main AnnData object" help="If checked, the main AnnData object will be outputted as well."/>
      </when>
      <when value="false"/>
    </conditional>
  </inputs>

  <outputs>
    <data name="output_h5ad" format="h5ad" from_work_dir="output.h5" label="${tool.name} on ${on_string}: @DESCRIPTION@ AnnData">
      <filter>output_format == 'anndata_h5ad' and (('output_main' in split_on_obs and split_on_obs['output_main']) or (not split_on_obs['default']))</filter>
    </data>
    <data name="output_h5" format="h5" from_work_dir="output.h5" label="${tool.name} on ${on_string}: @DESCRIPTION@ AnnData (h5)">
      <filter>output_format == 'anndata' and (('output_main' in split_on_obs and split_on_obs['output_main']) or (not split_on_obs['default']))</filter>
    </data>
    <collection name="output_h5ad_split" type="list" label="${tool.name} on ${on_string}: @DESCRIPTION@ AnnData split">
      <discover_datasets pattern="(?P&lt;designation&gt;.+)\.h5" directory="output_split" format="h5ad" visible="true"/>
      <filter>split_on_obs['default']</filter>
    </collection>
  </outputs>

  <tests>
    <test>
      <param name="input_obj_file" value="find_cluster.h5"/>
      <output name="output_h5ad" file="anndata_ops.h5" ftype="h5ad" compare="sim_size"/>
    </test>
    <test>
      <param name="input_obj_file" value="mnn.h5"/>
      <conditional name="swap_layer_to_x">
        <param name="default" value="true"/>
        <param name="layer" value="mnn"/>
        <param name="new_name_x" value="X_old"/>
      </conditional>
      <output name="output_h5ad" ftype="h5ad">
        <assert_contents>
          <has_h5_keys keys="layers/X_old" />
        </assert_contents>
      </output>
    </test>
    <test>
      <param name="input_obj_file" value="anndata_ops.h5"/>
      <param name="from_var" value = "gene_symbols" />
      <param name="to_var" value = "hello_all" />

      <output name="output_h5ad" file="anndata_ops_layer.h5" ftype="h5ad" compare="sim_size">
        <assert_contents>
          <has_h5_keys keys="layers/filtered" />
        </assert_contents>
      </output>
    </test>
    <test>
      <param name="input_obj_file" value="find_cluster.h5"/>
      <conditional name="split_on_obs">
        <param name="default" value="true"/>
        <param name="key" value="louvain"/>
        <param name="output_main" value="true"/>
      </conditional>
      <output name="output_h5ad" ftype="h5ad">
        <assert_contents>
          <has_h5_keys keys="obs/louvain" />
        </assert_contents>
      </output>
      <output_collection name="output_h5ad_split" type="list" count="5">
        <element name="louvain_0" ftype="h5ad">
          <assert_contents>
            <has_h5_keys keys="obs/louvain" />
          </assert_contents>
        </element>
      </output_collection>
    </test>
  </tests>

  <help><![CDATA[
=============================