Add 'extend' functionality

README.md

@@ -75,6 +75,16 @@ $1$ New organisms can be made available for download by adding them to the colle
 $2$ You can make your own primer3 settings by creating new or alterating existing JSON files in the `settings/primer3` folder. 
 
 
+## Extending previously-designed panels
+
+It's also possible to ask `multiply` to find designs that extend an existing multiplex primer panel (for example, one that has already been tested in the lab). To use this functionality:
+
+- specify the path to the `multiply` multiplex output for the previous designs - and a list of regions - in the `[Extend]` section of your design file.
+- then run the pipeline as usual.
+
+Multiple will search for primers in your regions of interest at the `generate` step, but will combine them with the previously-designed primers for all subsequent steps of the pipeline to build new multiplexes extending the original one. (An example of this process can be found in the `designs/pf-extend.ini` file.)
+
+
 ## Resources
 `multiply` uses the following external software and databases:
 - `primer3`. Individual primer pair design. https://primer3.org/

designs/pf-extend.ini

@@ -0,0 +1,25 @@
+[Sample]
+genome = PlasmodiumFalciparum
+
+[Genes]
+target_ids = PF3D7_0629500, PF3D7_0304600
+target_names = AAT1, CSP
+
+[Extend]
+target_ids = PF3D7_0206800, PF3D7_0810800, PF3D7_1407900, PF3D7_1408100, PF3D7_0709000, PF3D7_1343700, PF3D7_0417200, MDR1part
+target_names = MSP2, DHPS, PMI, PMIII, CRT1, K13, DHFR, MDR1part
+design_file = results/pf-default/select/table.multiplexes_information.csv
+
+[Primers]
+include_tails = False
+F_tail = TTTCTGTTGGTGCTGATATTGC
+R_tail = ACTTGCCTGTCGCTCTATCTTC
+
+[Amplicons]
+min_size_bp = 3000
+max_size_bp = 5000
+primer3_settings = default, stringent, relaxed, variable
+
+[Output]
+name = pf-extend
+primer_code = v

src/multiply/blast/annotator.py

@@ -82,7 +82,7 @@ def summarise_by_primer(self, output_path=None):
                 primer_pair_name=qseqid[:-2],
                 target_name=qseqid.split("_")[0],
                 total_alignments=qseqid_df.shape[0],
-                **qseqid_df[self.annotations].sum().to_dict(),
+                **qseqid_df[self.annotations.keys()].sum().to_dict(),
             )
             for qseqid, qseqid_df in self.blast_df.groupby("qseqid")
         ]

src/multiply/generate/main.py

@@ -140,7 +140,6 @@ def generate(design):
         print("Done.\n")
 
     # WRITE
-    print("Writing output table...")
     primer_df = pd.DataFrame(
         [
             pair.get_primer_as_dict(direction)
@@ -166,7 +165,16 @@ def generate(design):
             "pair_penalty",
         ]
     ]
+
+    if params['from_extend']:
+        print( f"Adding {params['extend_primers'].shape[0]} previously-generated primers from [Extend] section..." )
+        primer_df = pd.concat([
+            primer_df,
+            params['extend_primers']
+        ])
+
     output_csv = f"{params['output_dir']}/table.candidate_primers.csv"
+    print( f"Writing output table to \"{output_csv}\"..." )
     primer_df.to_csv(output_csv, index=False)
     print(f"  to: {output_csv}")
     print("Done.\n")

src/multiply/util/parsing.py

@@ -25,6 +25,7 @@ def parse_parameters(design_path):
     params = {}
     params = add_samples(config, params)
     params = add_genes(config, params)
+    params = add_extend(config, params)
     params = add_regions(config, params)
     params = add_primers(config, params)
     params = add_amplicons(config, params)
@@ -58,7 +59,7 @@ def check_design_exists(design_path):
 def check_valid_sections(
     config,
     must_include=["Sample", "Primers", "Amplicons", "Output"],
-    one_of=[["Genes", "Regions"]],
+    one_of=[["Genes", "Regions"]]
 ):
     """
     Check that all expected sections are found within the configuration object
@@ -91,7 +92,6 @@ def check_valid_sections(
                 f"Design must include at least one of these sections: {', '.join(section_set)}. Please add."
             )
 
-
 def add_samples(config, params):
     """
     Add [Sample] information to a parameter dictionary
@@ -174,6 +174,122 @@ def add_genes(config, params):
 
     return params
 
+def add_extend(config, params):
+    """
+    Add [Extend] information from a configparser object to a params dictionary
+    This is similar to the [Genes] section but lists previously-defined primers, via
+    a well-formed candidate primers table.
+    The input design_file can be, e.g. either the table.candidate_primers.csv or the
+    output table.multiplexes_information.csv from a previous run - it must have these columns:
+        'target_id',
+        'target_name',
+        'pair_name',
+        'primer_name',
+        'direction',
+        'seq',
+        'length',
+        'tm',
+        'gc',
+        'chrom',
+        'start',
+        'product_bp',
+        'pair_penalty'
+
+    params
+        config: ConfigParser
+            ConfigParser object holding design file information.
+        params: dict
+            Dictionary of MULTIPLY parameters.
+    returns
+        params: dict
+            Dictionary of MULTIPLY parameters, with [Extend]
+            parameters added.
+    """
+
+    # Check if genes have been provided
+    if not config.has_section("Extend"):
+        params["from_extend"] = False
+        return params
+
+    # Parse gene IDs
+    target_ids = [g.strip() for g in config.get("Extend", "target_ids").split(",")]
+
+    # Parse gene namess
+    has_names = config.has_option("Extend", "target_names")
+    if has_names:
+        target_names = [
+            g.strip() for g in config.get("Extend", "target_names").split(",")
+        ]
+    else:
+        target_names = target_ids
+
+    # Sanity checks
+    n_ids = len(target_ids)
+    n_names = len(target_names)
+    if not n_ids == n_names:
+        raise DesignFileError(
+            f"In [Extend], found {n_ids} `target_ids` and {n_names} `target_names`. Ensure equal."
+        )
+
+    # Load the previously-defined primers:
+    import pandas
+    design_file = config.get("Extend", "design_file" )
+    primer_df = pandas.read_csv( design_file )
+    primer_df = primer_df[
+        [
+            'target_id',
+            'target_name',
+            'pair_name',
+            'primer_name',
+            'direction',
+            'seq',
+            'length',
+            'tm',
+            'gc',
+            'chrom',
+            'start',
+            'product_bp',
+            'pair_penalty'
+        ]
+    ]
+    # Check all specified genes are represented in the file and have the correct names:
+    for index, id in enumerate( target_ids ):
+        this_target = primer_df[ primer_df.target_id == id ]
+        if this_target.shape[0] == 0:
+            raise DesignFileError(
+                f"In [Extend], target id \"{id}\" was not among the targets in the specified design file \"{design_file}\"."
+            )
+        target_name = list( set( this_target.target_name.to_list() ))
+        if len(target_name) > 1:
+            raise DesignFileError(
+                f"In [Extend], in the specified design file \"{design_file}\", target {target_id} seems to have more than one name."
+            )
+        target_name = target_name[0]
+        if has_names:
+            # If we're given names, make sure they are correct:
+            if target_name != target_names[index]:
+                raise DesignFileError(
+                    f"In [Extend], the name (\"{target_name}\") for target {id} in the design file \"{design_file}\", does not match the one specified (\"{target_names[index]}\")."
+                )
+        else:
+            # Otherwise get the name from the file:
+            target_names[index] = target_name
+
+    primer_df = primer_df[ primer_df.target_id.isin( target_ids )]
+
+    # Mapping between IDs and names
+    id_to_name = {i: n for i, n in zip(target_ids, target_names)}
+    name_to_id = {n: i for i, n in id_to_name.items()}
+
+    # Add to dictionary
+    params["from_extend"]       = True
+    params["extend_ids"]        = target_ids
+    params["extend_has_names"]  = True
+    params["extend_names"]      = target_names
+    params["extend_id_to_name"] = id_to_name
+    params["extend_primers"]    = primer_df
+
+    return params
 
 def add_regions(config, params):
     """