seer-pas-sdk 1.2.0__py3-none-any.whl → 1.2.2__py3-none-any.whl

seer_pas_sdk/common/__init__.py

@@ -99,7 +99,7 @@ def dict_to_df(data):
 
 
 # Most cases appear to be a .tsv file.
-def download_df(url, is_tsv=True, dtype={}):
+def download_df(url, is_tsv=True, dtype={}, usecols=None):
     """
     Fetches a TSV/CSV file from a URL and returns as a Pandas DataFrame.
 
@@ -114,6 +114,9 @@ def download_df(url, is_tsv=True, dtype={}):
     dtype : dict
         Data type conversion when intaking columns. e.g. {'a': str, 'b': np.float64}
 
+    usecols : list
+        Subset of columns to download. If not specified, downloads all columns.
+
     Returns
     -------
     pandas.core.frame.DataFrame
@@ -139,12 +142,10 @@ def download_df(url, is_tsv=True, dtype={}):
 
     if not url:
         return pd.DataFrame()
-    url_content = io.StringIO(requests.get(url).content.decode("utf-8"))
-    if is_tsv:
-        csv = pd.read_csv(url_content, sep="\t", dtype=dtype)
-    else:
-        csv = pd.read_csv(url_content, dtype=dtype)
-    return csv
+    csv = pd.read_csv(
+        url, sep="\t" if is_tsv else ",", usecols=usecols, engine="pyarrow"
+    )
+    return csv.astype(dtype=dtype) if dtype else csv
 
 
 def get_sample_info(

seer_pas_sdk/core/sdk.py

@@ -7,7 +7,6 @@ import requests
 import urllib.request
 import ssl
 
-
 from typing import List as _List, Tuple as _Tuple
 
 from ..common import *
@@ -15,8 +14,6 @@ from ..auth import Auth
 from ..objects.volcanoplot import VolcanoPlotBuilder
 from ..objects.headers import *
 
-import warnings
-
 
 class SeerSDK:
     """
@@ -1228,8 +1225,8 @@ class SeerSDK:
 
         >>> seer_sdk.get_msruns(sample_ids)
         >>> [
-            {"id": "SAMPLE_ID_1_HERE" ... },
-            {"id": "SAMPLE_ID_2_HERE" ... }
+            {"id": "MSRUN_ID_1_HERE" ... },
+            {"id": "MSRUN_ID_2_HERE" ... }
         ]
 
         >>> seer_sdk.get_msruns(sample_ids, as_df=True)
@@ -1295,8 +1292,8 @@ class SeerSDK:
 
         >>> seer_sdk.find_msruns(sample_ids)
         >>> [
-            {"id": "SAMPLE_ID_1_HERE" ... },
-            {"id": "SAMPLE_ID_2_HERE" ... }
+            {"id": "MSRUN_ID_1_HERE" ... },
+            {"id": "MSRUN_ID_2_HERE" ... }
         ]
 
         >>> seer_sdk.find_msruns(sample_ids, as_df=True)
@@ -1310,25 +1307,34 @@ class SeerSDK:
         URL = f"{self._auth.url}api/v1/msdatas/items"
 
         res = []
-        for sample_id in sample_ids:
 
-            with self._get_auth_session("findmsdatas") as s:
+        params = {"all": "true"}
 
-                msdatas = s.post(URL, json={"sampleId": sample_id})
+        with self._get_auth_session("findmsdatas") as s:
 
-                if msdatas.status_code != 200 or not msdatas.json()["data"]:
-                    raise ValueError(
-                        f"Failed to fetch MS data for sample ID={sample_id}."
-                    )
+            msdatas = s.post(
+                URL, json={"sampleId": ",".join(sample_ids)}, params=params
+            )
 
-                res += [x for x in msdatas.json()["data"]]
+            if msdatas.status_code != 200 or not msdatas.json()["data"]:
+                raise ValueError(
+                    f"Failed to fetch MS data for sample IDs={sample_ids}."
+                )
+
+            res += [x for x in msdatas.json()["data"]]
 
         spaces = {x["id"]: x["usergroup_name"] for x in self.get_spaces()}
+
+        def filepath_to_msrunid(filepath):
+            return os.path.basename(filepath).split(".")[0]
+
         for entry in res:
             if "tenant_id" in entry:
                 del entry["tenant_id"]
 
             if "raw_file_path" in entry:
+                # Provide a human-readable MS run id
+                entry["Run"] = filepath_to_msrunid(entry["raw_file_path"])
                 # Simple lambda function to find the third occurrence of '/' in the raw file path
                 location = lambda s: len(s) - len(s.split("/", 3)[-1])
                 # Slicing the string from the location
@@ -1339,6 +1345,13 @@ class SeerSDK:
                 entry["space"] = spaces.get(entry["user_group"], "General")
                 del entry["user_group"]
 
+            # Rename the key sample_id to sample_uuid
+            if "sample_id" in entry:
+                entry["sample_uuid"] = entry.pop("sample_id")
+            # Rename the key sample_id_tracking to sample_id
+            if "sample_id_tracking" in entry:
+                entry["sample_id"] = entry.pop("sample_id_tracking")
+
         if not res and as_df:
             return pd.DataFrame(columns=MSRUN_COLUMNS)
         return res if not as_df else dict_to_df(res)
@@ -1853,7 +1866,7 @@ class SeerSDK:
                             )
                         )
                     except Exception as e:
-                        print("Warning: Could not fetch fasta files.")
+                        print("Error: Could not fetch fasta files.")
                         res["fasta"] = None
                 else:
                     res["fasta"] = None
@@ -2066,7 +2079,7 @@ class SeerSDK:
                             )
                         except:
                             print(
-                                f"Warning: Could not fetch fasta files for analysis {res[entry].get('analysis_name')}."
+                                f"Error: Could not fetch fasta files for analysis {res[entry].get('analysis_name')}."
                             )
                 else:
                     res[entry]["fasta"] = None
@@ -2382,7 +2395,11 @@ class SeerSDK:
             return files
 
     def get_search_result(
-        self, analysis_id: str, analyte_type: str, rollup: str
+        self,
+        analysis_id: str,
+        analyte_type: str,
+        rollup: str,
+        columns: _List[str] = None,
     ):
         """
         Load one of the files available via the "Download result files" button on the PAS UI.
@@ -2423,6 +2440,7 @@ class SeerSDK:
                         "npLink"
                     ]["url"],
                     dtype=dtype,
+                    usecols=columns,
                 )
             elif rollup == "panel":
                 return download_df(
@@ -2430,6 +2448,7 @@ class SeerSDK:
                         "panelLink"
                     ]["url"],
                     dtype=dtype,
+                    usecols=columns,
                 )
         elif analyte_type == "peptide":
             if rollup == "np":
@@ -2438,6 +2457,7 @@ class SeerSDK:
                         "npLink"
                     ]["url"],
                     dtype=dtype,
+                    usecols=columns,
                 )
             elif rollup == "panel":
                 return download_df(
@@ -2445,12 +2465,14 @@ class SeerSDK:
                         "panelLink"
                     ]["url"],
                     dtype=dtype,
+                    usecols=columns,
                 )
         else:
             return download_df(
                 self.get_search_result_file_url(
                     analysis_id, filename="report.tsv"
-                )["url"]
+                )["url"],
+                usecols=columns,
             )
 
     def download_search_output_file(

seer_pas_sdk/core/unsupported.py

@@ -1471,37 +1471,70 @@ class _UnsupportedSDK(_SeerSDK):
         Get analyte intensities data for a given PAS analysis.
         Args:
             analysis_id (str): ID of the analysis.
-            analyte_type (str): Type of the analyte. Must be either 'protein', 'peptide', precursor.
+            analyte_type (str): Type of the analyte. Must be either 'protein', 'peptide', 'precursor'.
             rollup (str): Intensities rollup method. Must be either 'np' or 'panel'.
-            norm_method (str): Search engine. Supported engines are: raw, engine, median, median80, pepcal. Default is 'pepcal'.
+            norm_method (str): Search engine. Supported engines are: raw, engine, median, median80, pepcal, pepcal_batch. Default is 'pepcal'.
 
         Returns:
             pd.DataFrame: A dataframe with each row containing the analyte intensity measurement:
                             'msrun_id', 'sample_id', 'nanoparticle' (if rollup is 'np'), 'protein_group', 'peptide' (for 'peptide' and 'precursor' analyte types), 'charge' (for 'precursor' analyte type),
                             'intensity_log10', 'protein_group_q_value', 'q_value' (for 'precursor' analyte type), 'rt' and 'irt' (for 'peptide' and 'precursor' analyte types)
         """
-        # 1. Get msrun data for analysis
+
+        def filepath_to_msrunid(filepath):
+            return os.path.basename(filepath).split(".")[0]
+
+        # 1. Get samples and msrun data for analysis
         samples = self.find_samples(analysis_id=analysis_id)
-        sample_name_to_id = {s["sample_name"]: s["id"] for s in samples}
+
         sample_uuid_to_id = {s["id"]: s["sample_id"] for s in samples}
-        # for np rollup, a row represents an msrun
-        msruns = self.find_msruns(sample_ids=sample_name_to_id.values())
-        file_to_msrun = {
-            os.path.basename(msrun["raw_file_path"]).split(".")[0]: msrun
-            for msrun in msruns
-        }
-        sample_to_msrun = {msrun["sample_id"]: msrun for msrun in msruns}
+        sample_id_to_uuid = {s["sample_id"]: s["id"] for s in samples}
+        # FIXME sample_name is not guaranteed to be unique (within PAS analysis)
+        sample_name_to_uuid = {s["sample_name"]: s["id"] for s in samples}
 
-        # for panel rollup, a row represents a sample
+        msruns = self.find_msruns(sample_ids=[s["id"] for s in samples])
+        msrunid_to_info = {msrun["Run"]: msrun for msrun in msruns}
 
         # 2. Get search results
-        # pull the np/panel file, or report.tsv for precursor mode
+        # pull the np/panel file, or the relevant columns from the report.tsv for precursor mode
+        columns = None
+        if analyte_type == "precursor" and rollup == "np":
+            columnsExperiment = ["Run"]
+            columnsProtein = [
+                "Protein.Group",
+            ]
+            columnsPeptide = [
+                "Stripped.Sequence",
+            ]
+            columnsPrecursor = [
+                "Precursor.Id",
+                "Precursor.Charge",
+                "Precursor.Quantity",
+                "RT",
+                "iRT",
+                "IM",
+                "iIM",
+            ]
+            columnsQValue = [
+                "Q.Value",
+                "Protein.Q.Value",
+            ]
+            columns = [
+                *columnsExperiment,
+                *columnsProtein,
+                *columnsPeptide,
+                *columnsPrecursor,
+                *columnsQValue,
+            ]
         search_results = self.get_search_result(
             analysis_id=analysis_id,
             analyte_type=analyte_type,
             rollup=rollup,
+            columns=columns,
         )
+
         if analyte_type in ["protein", "peptide"]:
+            # set the intensity column based on norm_method and PAS analysis protocol version
             intensity_column = None
             if norm_method == "raw":
                 intensity_column = (
@@ -1543,139 +1576,171 @@ class _UnsupportedSDK(_SeerSDK):
                     raise ValueError(
                         "Pepcal normalized intensities not found in search results. This is only available with analyses processed with DIA-NN Seer Protocol v2.0 or later with the Seer Peptide Calibrant option enabled. \n Please retry using different norm_method, such as 'median'"
                     )
-
                 intensity_column = "PepCal Intensities Log10"
-
+            elif norm_method == "pepcal_batch":
+                if not (
+                    "PepCal Batch Intensities Log10" in search_results.columns
+                ):
+                    raise ValueError(
+                        "Pepcal normalized batch corrected intensities not found in search results. This is only available with analyses processed with DIA-NN Seer Protocol v2.0 or later with the Seer Peptide Calibrant option enabled. \n Please retry using different norm_method, such as 'median'"
+                    )
+                intensity_column = "PepCal Batch Intensities Log10"
             else:
                 raise ValueError(
                     f"norm_method = {norm_method} is not supported. Supported normalization methods are: raw, pepcal, engine, median, median80."
                 )
-            if rollup == "panel":
-                search_results.fillna({"Sample Name": ""}, inplace=True)
-                search_results["File Name"] = search_results[
-                    "Sample Name"
-                ].apply(
-                    lambda x: (
-                        os.path.basename(
-                            sample_to_msrun[sample_name_to_id[x]][
-                                "raw_file_path"
-                            ]
-                        ).split(".")[0]
-                        if x
-                        else None
-                    )
-                )
-            search_results["File Name"] = search_results["File Name"].apply(
-                lambda x: os.path.basename(x).split(".")[0] if x else None
-            )
 
             search_results["Intensity Log10"] = search_results[
                 intensity_column
             ]
 
-            # 3. Merge report to search results to get Q value and other properties
-            report = self.get_search_result(
-                analysis_id=analysis_id,
-                analyte_type="precursor",
-                rollup="np",
-            )
-            report["File Name"] = report["Run"]
-            report["Protein Group"] = report["Protein.Group"]
-
-            if analyte_type == "protein":
-                report["Protein Q Value"] = report["Protein.Q.Value"]
-
-                report = report[
-                    ["File Name", "Protein Group", "Protein Q Value"]
-                ]
-                report.drop_duplicates(
-                    subset=["File Name", "Protein Group"], inplace=True
+            if rollup == "panel":
+                search_results.rename(
+                    columns={"Sample ID": "Sample UUID"}, inplace=True
                 )
-                df = pd.merge(
-                    search_results,
-                    report,
-                    on=["File Name", "Protein Group"],
-                    how="left",
+                search_results["Sample UUID"] = search_results[
+                    "Sample Name"
+                ].map(sample_name_to_uuid)
+                search_results["Sample ID"] = search_results[
+                    "Sample UUID"
+                ].map(sample_uuid_to_id)
+                experiment_columns = ["Sample UUID", "Sample ID"]
+
+                # analyte info is limited to the id in the panel rollup
+                if analyte_type == "protein":
+                    analyte_id_column = "Protein Group"
+                else:
+                    analyte_id_column = "Peptide"
+
+                analyte_columns = [analyte_id_column]
+                df = search_results
+            else:
+                # np rollup, extract basename without extension
+                path_to_msrunid = {
+                    path: filepath_to_msrunid(path)
+                    for path in search_results["File Name"].unique()
+                }
+                # strip path from the filename to allow merging with the precursor report
+                search_results["Run"] = search_results["File Name"].map(
+                    path_to_msrunid
                 )
-                included_columns = [
-                    "MsRun ID",
-                    "Sample ID",
-                    "Protein Group",
-                    "Intensity Log10",
-                    "Protein Q Value",
-                ]
 
-            else:
-                report["Peptide"] = report["Stripped.Sequence"]
-                #  If analyte_type is peptide, attach retention time (RT, iRT)
-                report = report[["File Name", "Peptide", "RT", "iRT"]]
-                report.drop_duplicates(
-                    subset=["File Name", "Peptide"], inplace=True
+                search_results["MsRun UUID"] = search_results["Run"].map(
+                    {k: v["id"] for k, v in msrunid_to_info.items()}
                 )
-                df = pd.merge(
-                    search_results,
-                    report,
-                    on=["File Name", "Peptide"],
-                    how="left",
+                search_results["Sample ID"] = search_results["Run"].map(
+                    {k: v["sample_id"] for k, v in msrunid_to_info.items()}
+                )
+                search_results["Sample UUID"] = search_results["Run"].map(
+                    {k: v["sample_uuid"] for k, v in msrunid_to_info.items()}
                 )
-                included_columns = [
-                    "MsRun ID",
+                search_results["Nanoparticle"] = search_results["Run"].map(
+                    {k: v["nanoparticle"] for k, v in msrunid_to_info.items()}
+                )
+                experiment_columns = [
+                    "MsRun UUID",
+                    "Run",
+                    "Nanoparticle",
+                    "Sample UUID",
                     "Sample ID",
-                    "Peptide",
-                    "Protein Group",
-                    "Intensity Log10",
-                    "RT",
-                    "iRT",
                 ]
-            # endif
 
-            if rollup == "np":
-                included_columns.insert(
-                    included_columns.index("Sample ID") + 1, "Nanoparticle"
+                # Merge report to search results to get Q value and other properties
+                if analyte_type == "protein":
+                    columns = ["Run", "Protein.Group", "Protein.Q.Value"]
+                elif analyte_type == "peptide":
+                    columns = ["Run", "Stripped.Sequence", "Protein.Q.Value"]
+                analytes = self.get_search_result(
+                    analysis_id=analysis_id,
+                    analyte_type="precursor",
+                    rollup="np",
+                    columns=columns,
+                )
+                # pandas Dataframe.rename() default behavior is to ignore the columns that do not exist in the data frame.
+                analytes.rename(
+                    columns={
+                        "Protein.Group": "Protein Group",
+                        "Protein.Q.Value": "Protein Q Value",
+                        "Stripped.Sequence": "Peptide",
+                    },
+                    inplace=True,
                 )
 
-            df["MsRun ID"] = df["File Name"].apply(
-                lambda x: (
-                    file_to_msrun[x]["id"] if x in file_to_msrun else None
+                if analyte_type == "protein":
+                    analyte_id_column = "Protein Group"
+                    analyte_columns = [
+                        analyte_id_column,
+                        "Protein Q Value",
+                    ]
+
+                else:
+                    analyte_id_column = "Peptide"
+                    analyte_columns = [analyte_id_column]
+                # endif analyte_type
+
+                analytes.drop(
+                    columns=[
+                        col
+                        for col in analytes.columns
+                        if col != "Run" and col not in analyte_columns
+                    ],
+                    inplace=True,
                 )
-            )
-            df["Sample ID"] = df["File Name"].apply(
-                lambda x: (
-                    file_to_msrun[x]["sample_id"]
-                    if x in file_to_msrun
-                    else None
+                analytes.drop_duplicates(
+                    subset=["Run", analyte_id_column], inplace=True
                 )
-            )
-            df = df[included_columns]
+                df = pd.merge(
+                    search_results,
+                    analytes,
+                    on=["Run", analyte_id_column],
+                    how="left",
+                    validate="one_to_one",
+                )
+
+            df = df[experiment_columns + analyte_columns + ["Intensity Log10"]]
 
         else:
             # precursor
             # working only in report.tsv
-            search_results["Intensity"] = search_results["Precursor.Quantity"]
-            search_results["MsRun ID"] = search_results["Run"].apply(
-                lambda x: (
-                    file_to_msrun[x]["id"] if x in file_to_msrun else None
+            if norm_method != "raw":
+                raise ValueError(
+                    "For precursor analyte type, only 'raw' norm_method is supported."
                 )
+
+            search_results["MsRun UUID"] = search_results["Run"].map(
+                {k: v["id"] for k, v in msrunid_to_info.items()}
             )
-            search_results["Sample ID"] = search_results["Run"].apply(
-                lambda x: (
-                    file_to_msrun[x]["sample_id"]
-                    if x in file_to_msrun
-                    else None
-                )
+            search_results["Sample ID"] = search_results["Run"].map(
+                {k: v["sample_id"] for k, v in msrunid_to_info.items()}
             )
-            search_results["Protein Group"] = search_results["Protein.Group"]
-            search_results["Peptide"] = search_results["Stripped.Sequence"]
-            search_results["Charge"] = search_results["Precursor.Charge"]
-            search_results["Precursor Id"] = search_results["Precursor.Id"]
-            search_results["Precursor Q Value"] = search_results["Q.Value"]
-            search_results["Protein Q Value"] = search_results[
-                "Protein.Q.Value"
+            search_results["Sample UUID"] = search_results["Sample ID"].map(
+                sample_id_to_uuid
+            )
+            search_results["Nanoparticle"] = search_results["Run"].map(
+                {k: v["nanoparticle"] for k, v in msrunid_to_info.items()}
+            )
+            experiment_columns = [
+                "MsRun UUID",
+                "Run",
+                "Nanoparticle",
+                "Sample UUID",
+                "Sample ID",
             ]
 
-            included_columns = [
-                "MsRun ID",
-                "Sample ID",
+            search_results.rename(
+                columns={
+                    "Protein.Group": "Protein Group",
+                    "Stripped.Sequence": "Peptide",
+                    "Precursor.Charge": "Charge",
+                    "Precursor.Id": "Precursor Id",
+                    "Q.Value": "Precursor Q Value",
+                    "Protein.Q.Value": "Protein Q Value",
+                    "Precursor.Quantity": "Intensity",
+                },
+                inplace=True,
+            )
+
+            analyte_columns = [
                 "Protein Group",
                 "Protein Q Value",
                 "Peptide",
@@ -1688,16 +1753,12 @@ class _UnsupportedSDK(_SeerSDK):
                 "IM",
                 "iIM",
             ]
-            df = pd.DataFrame(search_results[included_columns])
+            df = pd.DataFrame(
+                search_results[experiment_columns + analyte_columns]
+            )
 
         df.columns = [title_case_to_snake_case(x) for x in df.columns]
-        df["sample_uuid"] = df["sample_id"]
-        df["sample_id"] = df["sample_uuid"].apply(
-            lambda x: sample_uuid_to_id.get(x)
-        )
 
-        if rollup == "panel":
-            df.drop(columns=["msrun_id"], inplace=True, errors="ignore")
         return df
 
     def get_search_data_analytes(self, analysis_id: str, analyte_type: str):
@@ -1714,10 +1775,6 @@ class _UnsupportedSDK(_SeerSDK):
             analysis_id=analysis_id, analyte_type="protein", rollup="np"
         )
 
-        report_results = self.get_search_result(
-            analysis_id=analysis_id, analyte_type="precursor", rollup="np"
-        )
-
         search_results = search_results[
             [
                 "Protein Group",
@@ -1729,18 +1786,87 @@ class _UnsupportedSDK(_SeerSDK):
             ]
         ]
         search_results.drop_duplicates(subset=["Protein Group"], inplace=True)
-        report_results["Protein Group"] = report_results["Protein.Group"]
-        report_results["Peptide"] = report_results["Stripped.Sequence"]
 
-        if analyte_type == "protein":
-            report_results = report_results[
+        # 2. fetch precursor report to extract analyte-specific details
+        columnsPG = [
+            "Protein.Group",
+        ]
+        columnsPeptide = [
+            "Protein.Ids",
+            "Stripped.Sequence",
+            "Proteotypic",
+        ]
+        columnsPrecursor = [
+            "Precursor.Id",
+            "Precursor.Charge",
+            "Precursor.Quantity",
+            "Modified.Sequence",
+        ]
+        columnsPGQValue = [
+            "Global.PG.Q.Value",
+            "Lib.PG.Q.Value",
+        ]
+        columnsPrecursorQValue = [
+            "Global.Q.Value",
+            "Lib.Q.Value",
+        ]
+        columns = [
+            *columnsPG,
+            *columnsPGQValue,
+        ]
+        if analyte_type == "peptide":
+            columns += [*columnsPeptide]
+        elif analyte_type == "precursor":
+            columns += [
+                *columnsPeptide,
+                *columnsPrecursor,
+                *columnsPrecursorQValue,
+            ]
+        report_results = self.get_search_result(
+            analysis_id=analysis_id,
+            analyte_type="precursor",
+            rollup="np",
+            columns=columns,
+        )
+        report_results.rename(
+            columns={
+                "Protein.Group": "Protein Group",
+                "Stripped.Sequence": "Peptide",
+                "Modified.Sequence": "Modified.Peptide",
+            },
+            inplace=True,
+        )
+
+        # function to fix the potential bug, where different precursors
+        # of the same peptide map to different protein groups
+        def fix_peptide_to_protein_group_assignment(
+            df: pd.DataFrame,
+        ) -> pd.DataFrame:
+            # for each peptide, sort protein groups by confidence
+            df = df.sort_values(
                 [
-                    "Protein Group",
-                    "Protein.Ids",
+                    "Peptide",
                     "Global.PG.Q.Value",
                     "Lib.PG.Q.Value",
+                    "Protein Group",
                 ]
-            ]
+            )
+
+            # broadcast the best protein group across all rows with the same peptide
+            # to fix the potential bug, where different precursors of the same peptide
+            # map to different protein groups
+            for col in [
+                "Protein Group",
+                "Protein.Ids",
+                "Protein.Names",
+                "Genes",
+            ]:
+                if col in df.columns:
+                    df[col] = df.groupby("Peptide")[col].transform("first")
+
+            return df
+
+        if analyte_type == "protein":
             report_results.drop_duplicates(
                 subset=["Protein Group"], inplace=True
             )
@@ -1751,41 +1877,18 @@ class _UnsupportedSDK(_SeerSDK):
                 how="left",
             )
         elif analyte_type == "peptide":
-
-            # The below logic performs the following:
-            # 1. orders each peptide group by Global.PG.Q.Value, Lib.PG.Q.Value, and Protein Group (ascending)
-            # 2. for each peptide group, select the first row to find the precursor with the lowest Q values
-            # 3. broadcasts the associated protein group columns across all rows with the same peptide.
-            #
-            # This ensures that for each peptide, we retain consistent protein information while avoiding duplication.
-
-            report_results = report_results.sort_values(
+            search_results = search_results[
                 [
-                    "Peptide",
-                    "Global.PG.Q.Value",
-                    "Lib.PG.Q.Value",
                     "Protein Group",
+                    "Protein Names",
+                    "Gene Names",
                 ]
+            ]
+            report_results.drop_duplicates(inplace=True)
+            report_results = fix_peptide_to_protein_group_assignment(
+                report_results
             )
-
-            columns_to_broadcast = ["Protein Group", "Protein.Ids"]
-            broadcasted = (
-                report_results.groupby("Peptide")
-                .apply(
-                    lambda x: pd.Series(
-                        {
-                            col: x.iloc[0][col]
-                            for col in columns_to_broadcast + ["Peptide"]
-                        }
-                    )
-                )
-                .reset_index(drop=True)
-            )
-            report_results = (
-                report_results.drop(columns=columns_to_broadcast)
-                .merge(broadcasted, on="Peptide", how="left")
-                .drop_duplicates(subset=["Peptide"])
-            )
+            report_results.drop_duplicates(subset=["Peptide"], inplace=True)
 
             df = pd.merge(
                 report_results,
@@ -1793,15 +1896,6 @@ class _UnsupportedSDK(_SeerSDK):
                 on=["Protein Group"],
                 how="left",
             )
-            df = df[
-                [
-                    "Peptide",
-                    "Protein Group",
-                    "Protein.Ids",
-                    "Protein Names",
-                    "Gene Names",
-                ]
-            ]
         else:
             # precursor
             search_results = search_results[
@@ -1811,91 +1905,23 @@ class _UnsupportedSDK(_SeerSDK):
                     "Gene Names",
                 ]
             ]
-            search_results.drop_duplicates(
-                subset=["Protein Group"], inplace=True
-            )
-            report_results = report_results[
-                [
-                    "Precursor.Id",
-                    "Precursor.Charge",
-                    "Peptide",
-                    "Protein Group",
-                    "Protein.Ids",
-                    "Protein.Names",
-                    "Genes",
-                    "Modified.Sequence",
-                    "Proteotypic",
-                    "Global.Q.Value",
-                    "Global.PG.Q.Value",
-                    "Lib.Q.Value",
-                    "Lib.PG.Q.Value",
-                ]
-            ]
+            report_results.drop_duplicates(inplace=True)
 
-            # The below logic performs the following:
-            # 1. orders each peptide group by Global.PG.Q.Value, Lib.PG.Q.Value, and Protein Group (ascending)
-            # 2. for each peptide group, select the first row to find the precursor with the lowest Q values
-            # 3. broadcasts the associated protein group columns across all rows with the same peptide.
-            #
-            # This ensures that for each peptide, we retain consistent protein information while avoiding duplication.
-            columns_to_broadcast = [
-                "Protein Group",
-                "Protein.Ids",
-                "Protein.Names",
-                "Genes",
-            ]
-            report_results = report_results.sort_values(
-                [
-                    "Peptide",
-                    "Global.PG.Q.Value",
-                    "Lib.PG.Q.Value",
-                    "Protein Group",
-                ],
-            )
-            broadcasted = (
-                report_results.groupby("Peptide")
-                .apply(
-                    lambda x: pd.Series(
-                        {
-                            col: x.iloc[0][col]
-                            for col in columns_to_broadcast + ["Peptide"]
-                        }
-                    )
-                )
-                .reset_index(drop=True)
+            report_results = fix_peptide_to_protein_group_assignment(
+                report_results
             )
-            report_results = (
-                report_results.drop(columns=columns_to_broadcast)
-                .merge(broadcasted, on="Peptide", how="left")
-                .drop_duplicates(subset=["Peptide", "Precursor.Charge"])
+            report_results.drop_duplicates(
+                subset=["Peptide", "Modified.Peptide", "Precursor.Charge"],
+                inplace=True,
             )
+
             df = pd.merge(
                 report_results,
                 search_results,
                 on=["Protein Group"],
                 how="left",
             )
-            df = df[
-                [
-                    "Precursor.Id",
-                    "Precursor.Charge",
-                    "Peptide",
-                    "Protein Group",
-                    "Protein.Ids",
-                    "Protein.Names",
-                    "Genes",
-                    "Modified.Sequence",
-                    "Proteotypic",
-                    "Global.Q.Value",
-                    "Global.PG.Q.Value",
-                    "Lib.Q.Value",
-                    "Lib.PG.Q.Value",
-                    "Gene Names",
-                ]
-            ]
-            df.rename(
-                columns={"Modified.Sequence": "Modified.Peptide"}, inplace=True
-            )
         # endif
         df.columns = [title_case_to_snake_case(x) for x in df.columns]
+
         return df

{seer_pas_sdk-1.2.0.dist-info → seer_pas_sdk-1.2.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: seer-pas-sdk
-Version: 1.2.0
+Version: 1.2.2
 Summary: SDK for Seer Proteograph Analysis Suite (PAS)
 Author-email: Ryan Sun <rsun@seer.bio>
 License: 
@@ -194,9 +194,10 @@ License-File: LICENSE.txt
 Requires-Dist: boto3>=1.26.152
 Requires-Dist: botocore>=1.29.152
 Requires-Dist: pandas>=2.0.1
+Requires-Dist: pyarrow>=17.0.0
 Requires-Dist: PyJWT>=2.8.0
 Requires-Dist: python-dotenv>=1.0.0
-Requires-Dist: Requests>=2.31.0
+Requires-Dist: requests>=2.31.0
 Requires-Dist: tqdm>=4.65.0
 Requires-Dist: deprecation
 Dynamic: license-file

{seer_pas_sdk-1.2.0.dist-info → seer_pas_sdk-1.2.2.dist-info}/RECORD

@@ -1,19 +1,19 @@
 seer_pas_sdk/__init__.py,sha256=Ie6atdmdBV-OmdHHXjhrGhdFGXiyP3JKhKrr3hyvSsA,563
 seer_pas_sdk/auth/__init__.py,sha256=e_eM4jJnnyKUdg4Nggzi9ypt2MLWcEJ8CmCPkUaQDSs,23
 seer_pas_sdk/auth/auth.py,sha256=_SI5CdEkfqfr4o5BQ79BuPbxGeI9p7tqxJd7mUqSAkI,8854
-seer_pas_sdk/common/__init__.py,sha256=LLfkbsZMXXty_T8xkOAws_WWBpbfwWZAdkNTduS8Abc,24443
+seer_pas_sdk/common/__init__.py,sha256=WrRwmSONUdFD0ysT5jHwG2zWDd-v2wverjXY7BWZhHU,24488
 seer_pas_sdk/common/errors.py,sha256=4HFORWnaQQCMXRE8kwdsJWvQRB_3KFEZ7yMb391e4gA,142
 seer_pas_sdk/common/groupanalysis.py,sha256=DxB-gbQfYzl7p9MTYWDIqghcH-IeakzdYdrRZrlIHek,1730
 seer_pas_sdk/core/__init__.py,sha256=rxbKgg-Qe24OaxX2zyHHYPYgDCTEKE_-41bB2wvpvL4,25
-seer_pas_sdk/core/sdk.py,sha256=yDml92xhZtWR54-MgWG3rYVVlcaaAl2i6EzlWgbit8Q,160705
-seer_pas_sdk/core/unsupported.py,sha256=SpxKQx_SN0o7SEBGXko_vmTQVxYDAvXEQGH2VWTK63M,71915
+seer_pas_sdk/core/sdk.py,sha256=0ukg287lsjlSNoV0WqFbiPMURhVogsy_sTR7gg1fr9Q,161512
+seer_pas_sdk/core/unsupported.py,sha256=WcF_Z6ZUpzOWkWQHaMtm9SnE2NveuRmljVfNe8QSbms,72732
 seer_pas_sdk/objects/__init__.py,sha256=r-lY7axLTzToAI-Dme019YfcJLDe2ok1f_e6OQx3j64,130
 seer_pas_sdk/objects/groupanalysis.py,sha256=x3D_5NmYBoPDilNCQqUoCFARIfIeUq4FBY3_N6u8tfM,994
 seer_pas_sdk/objects/headers.py,sha256=RilNzB_Nhid3U8j93BxJYcRrgDmd_1bAuI0P465xd0g,2727
 seer_pas_sdk/objects/platemap.py,sha256=8IvJPAecs_e_FyqibzhCw-O4zjCFnf-zMUp_5krTEsg,5864
 seer_pas_sdk/objects/volcanoplot.py,sha256=lTrTOVg74nT3uo-P1edQJC1ZbdoiLMtQ3VJd9CnzmoM,9396
-seer_pas_sdk-1.2.0.dist-info/licenses/LICENSE.txt,sha256=DVQuDIgE45qn836wDaWnYhSdxoLXgpRRKH4RuTjpRZQ,10174
-seer_pas_sdk-1.2.0.dist-info/METADATA,sha256=6mQ4VXcrUHCVfn3PL0pD2j_u7yJNMmw1HRqtd_lATDg,13413
-seer_pas_sdk-1.2.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-seer_pas_sdk-1.2.0.dist-info/top_level.txt,sha256=-2kZ-KFMGtXwr8H1O5llMKlcJ8gRKohEmrIvazXB61s,13
-seer_pas_sdk-1.2.0.dist-info/RECORD,,
+seer_pas_sdk-1.2.2.dist-info/licenses/LICENSE.txt,sha256=DVQuDIgE45qn836wDaWnYhSdxoLXgpRRKH4RuTjpRZQ,10174
+seer_pas_sdk-1.2.2.dist-info/METADATA,sha256=Lw-pb90n0vo7K4I-2wnQm_LvsrVzXpEYmj8t4vwToAA,13444
+seer_pas_sdk-1.2.2.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+seer_pas_sdk-1.2.2.dist-info/top_level.txt,sha256=-2kZ-KFMGtXwr8H1O5llMKlcJ8gRKohEmrIvazXB61s,13
+seer_pas_sdk-1.2.2.dist-info/RECORD,,

{seer_pas_sdk-1.2.0.dist-info → seer_pas_sdk-1.2.2.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.9.0)
+Generator: setuptools (80.10.2)
 Root-Is-Purelib: true
 Tag: py3-none-any