geney 1.1.10__py2.py3-none-any.whl → 1.1.11__py2.py3-none-any.whl

geney/immune_utils.py

@@ -0,0 +1,185 @@
+import subprocess
+import logging
+import tempfile
+from geney import config_setup
+import pandas as pd
+
+
+class NetChop(object):
+    """
+    Wrapper around netChop tool. Assumes netChop is in your PATH.
+    """
+    def predict_epitopes(self, sequences, threshold=0.5, min_len=8):
+        """
+        Return netChop predictions for each position in each sequence.
+
+        Parameters
+        -----------
+        sequences : list of string
+            Amino acid sequences to predict cleavage for
+
+        Returns
+        -----------
+        list of list of float
+
+        The i'th list corresponds to the i'th sequence. Each list gives
+        the cleavage probability for each position in the sequence.
+        """
+        with tempfile.NamedTemporaryFile(dir=config_setup['NETCHOP'], suffix=".fsa", mode="w") as input_fd:
+            for (i, sequence) in enumerate(sequences):
+                _ = input_fd.write("> %d\n" % i)
+                _ = input_fd.write(sequence)
+                _ = input_fd.write("\n")
+            input_fd.flush()
+            try:
+                output = subprocess.check_output(["netchop", str(input_fd.name)])
+            except subprocess.CalledProcessError as e:
+                logging.error("Error calling netChop: %s:\n%s" % (e, e.output))
+                raise
+        parsed = self.parse_netchop(output)
+        return parsed
+        #
+        # assert len(parsed) == len(sequences), \
+        #     "Expected %d results but got %d" % (
+        #         len(sequences), len(parsed))
+        # assert [len(x) for x in parsed] == [len(x) for x in sequences]
+        # filtered_proteosomes = []
+        # for scores, seq in list(zip(parsed, sequences)):
+        #     proteosome = self.chop_protein(seq, [s > threshold for s in scores])
+        #     filtered_proteosomes.append([e for e in proteosome if len(e) > min_len])
+        # return filtered_proteosomes
+    @staticmethod
+    def parse_netchop(netchop_output):
+        """
+        Parse netChop stdout.
+        """
+        line_iterator = iter(netchop_output.decode().split("\n"))
+        scores = []
+        for line in line_iterator:
+            if "pos" in line and 'AA' in line and 'score' in line:
+                scores.append([])
+                if "----" not in next(line_iterator):
+                    raise ValueError("Dashes expected")
+                line = next(line_iterator)
+                while '-------' not in line:
+                    score = float(line.split()[3])
+                    scores[-1].append(score)
+                    line = next(line_iterator)
+        return scores
+    def chop_protein(self, seq, pos):
+        # Generate subsequences using list comprehension and slicing
+        start = 0
+        subsequences = [seq[start:(start := i+1)] for i, marker in enumerate(pos) if marker == 1]
+        # Check if the last part needs to be added
+        if start < len(seq):
+            subsequences.append(seq[start:])
+        return subsequences
+    def generate_cut_sequences(self, char_sequence, cut_probabilities):
+        """
+        Generate all possible cut sequences and their abundance values,
+        considering only those sequences where the probabilities of all cut sites
+        between the two ends are zero.
+
+        :param char_sequence: A string representing the sequence of characters.
+        :param cut_probabilities: A list of probabilities for each position in the sequence.
+        :return: A list of tuples, where each tuple contains a cut sequence and its abundance value.
+        """
+        if len(char_sequence) != len(cut_probabilities):
+            raise ValueError("Character sequence and cut probabilities must have the same length.")
+        cut_sequences = []
+        # Generate all possible cuts
+        for i in range(len(char_sequence)):
+            for j in range(i + 1, len(char_sequence) + 1):
+                # Check if probabilities of all cut sites between i and j are zero
+                if sum(cut_probabilities[i + 1:j - 1]) < 1:
+                    cut_sequence = char_sequence[i:j]
+                    abundance_value = cut_probabilities[i] * cut_probabilities[j - 1] - sum(
+                        cut_probabilities[i + 1:j - 1])
+                    cut_sequences.append({'seq': cut_sequence, 'abundance': abundance_value})
+        return pd.DataFrame(cut_sequences)
+
+
+
+
+from .base_commandline_predictor import BaseCommandlinePredictor
+from .parsing import parse_netmhc41_stdout
+from functools import partial
+
+
+class NetMHCpan41(BaseCommandlinePredictor):
+    def __init__(
+            self,
+            alleles,
+            default_peptide_lengths=[9],
+            program_name="netMHCpan",
+            process_limit=-1,
+            mode="binding_affinity",
+            extra_flags=[]):
+        """
+        Wrapper for NetMHCpan4.1.
+
+        The mode argument should be one of "binding_affinity" (default) or
+        "elution_score".
+        """
+
+        # The -BA flag is required to predict binding affinity
+        if mode == "binding_affinity":
+            flags = ["-BA"]
+        elif mode == "elution_score":
+            flags = []
+        else:
+            raise ValueError("Unsupported mode", mode)
+
+        BaseCommandlinePredictor.__init__(
+            self,
+            program_name=program_name,
+            alleles=alleles,
+            default_peptide_lengths=default_peptide_lengths,
+            parse_output_fn=partial(parse_netmhc41_stdout, mode=mode),
+            supported_alleles_flag="-listMHC",
+            input_file_flag="-f",
+            length_flag="-l",
+            allele_flag="-a",
+            extra_flags=flags + extra_flags,
+            process_limit=process_limit)
+
+class NetMHCpan41_EL(NetMHCpan41):
+    """
+    Wrapper for NetMHCpan4 when the preferred mode is elution score
+    """
+    def __init__(
+            self,
+            alleles,
+            default_peptide_lengths=[9],
+            program_name="netMHCpan",
+            process_limit=-1,
+            extra_flags=[]):
+        NetMHCpan41.__init__(
+            self,
+            alleles=alleles,
+            default_peptide_lengths=default_peptide_lengths,
+            program_name=program_name,
+            process_limit=process_limit,
+            mode="elution_score",
+            extra_flags=extra_flags)
+
+
+class NetMHCpan41_BA(NetMHCpan41):
+    """
+    Wrapper for NetMHCpan4 when the preferred mode is binding affinity
+    """
+    def __init__(
+            self,
+            alleles,
+            default_peptide_lengths=[9],
+            program_name="netMHCpan",
+            process_limit=-1,
+            extra_flags=[]):
+        NetMHCpan41.__init__(
+            self,
+            alleles=alleles,
+            default_peptide_lengths=default_peptide_lengths,
+            program_name=program_name,
+            process_limit=process_limit,
+            mode="binding_affinity",
+            extra_flags=extra_flags)

geney/power_utils.py

@@ -65,6 +65,7 @@ def launch_dask_cluster(memory_size="3GB", num_workers=10, queue="tamirQ",
                 log_directory=log_directory,
                 job_script_prologue=[f"cd {config_setup['BASE']}"]
             )
+
         else:
             dask_cluster = PBSCluster(
                 cores=1,

geney/survival_utils.py

@@ -82,13 +82,15 @@ class SurvivalAnalysis:
 
             kmf.fit(g['duration'], g['event'], label=label)
             surv_func = kmf.survival_function_
-            auc = trapz(surv_func[label], surv_func.index)
+            filtered_surv_func = surv_func[surv_func.index <= cap_time]
+            auc = np.trapz(filtered_surv_func[label], filtered_surv_func.index)
+            # auc = trapz(surv_func[label], surv_func.index)
             auc_vals.append(auc)
             if plot:
                 if count == 0:
-                    ax = kmf.plot()
+                    ax = kmf.plot_survival_function()
                 else:
-                    kmf.plot(ax=ax)
+                    kmf.plot_survival_function(ax=ax)
                 count += 1
         p_value = self.log_rank(df[df[feature] == 1], df[df[feature] == 0])
 

geney/tcga_utils.py

@@ -157,12 +157,10 @@ class TCGAGene:
         # returns two lists: all patients affected by a mutation and all patients with none of the mutations (or the mutations but not togehter)
         pass
 
-
     def arrange_patients_by_project(self, mut_id):
         # returns all the patients affected by a mutation grouped by cancer project
         pass
 
-
     def total_prevalence(self, mut_id):
         pass
 

{geney-1.1.10.dist-info → geney-1.1.11.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: geney
-Version: 1.1.10
+Version: 1.1.11
 Summary: A Python package for gene expression modeling.
 Home-page: https://github.com/nicolaslynn/geney
 Author: Nicolas Lynn

{geney-1.1.10.dist-info → geney-1.1.11.dist-info}/RECORD

@@ -7,15 +7,16 @@ geney/config_setup.py,sha256=SePeooA4RWAtR_KAT1-W1hkD3MT5tH6YMyp80t_RNPQ,385
 geney/data_setup.py,sha256=DZeksRPr2ZT7bszMo33W0r3OwmqHokVXtZ4gx5Lu_Mo,10725
 geney/gtex.py,sha256=asL2lHyU5KsbWpV096vkf1Ka7hSo_RRfZqw7p5nERmE,1919
 geney/gtex_utils.py,sha256=asL2lHyU5KsbWpV096vkf1Ka7hSo_RRfZqw7p5nERmE,1919
+geney/immune_utils.py,sha256=0udmTxqF9jCYeUOgP7bGLWMEBH3KBikKu8pPQnE9Rfo,6881
 geney/netchop.py,sha256=AMiy9YsdTmX4B3k3Y5Yh7EmoGAojM1O3AzhPKOiB--g,3050
 geney/oncosplice.py,sha256=Fyc_UtAhV3Pv0vk8V55rO_jnb2Dwj5sW98KVwP3PHwU,68964
 geney/oncosplice_pipeline.py,sha256=hpGqFHOdn8i8tvvs1-t3-G9Ko18zInwoDXBJbbrfbC4,68036
 geney/performance_utils.py,sha256=FQt7rA4r-Wuq3kceCxsSuMfj3wU1tMG8QnbL59aBohs,4700
-geney/power_utils.py,sha256=WRpqMnqUv1xrAeTduAUhx6YpSEJQci7bC2od12JcVtE,7267
+geney/power_utils.py,sha256=6InuDm1jSrsgR-F_LmdMTbuQwty2OdYjwfGGaAPhaRI,7268
 geney/survival.py,sha256=gNKZGcwxDZ00ixVBHf3ZdjbY_AHQOCU9kKpBC_dokbM,5572
-geney/survival_utils.py,sha256=gNKZGcwxDZ00ixVBHf3ZdjbY_AHQOCU9kKpBC_dokbM,5572
+geney/survival_utils.py,sha256=2CAkC2LsspicHIdrqsiPnjgvpr5KHDUfLFFqnRbPJqs,5762
 geney/tcga_annotations.py,sha256=DjRl6Pk5VAOL1yhbt8SXD6FZhYbcYNu3FtXYMeveGB0,15016
-geney/tcga_utils.py,sha256=XrLI8RzmXhyabvL24sMrqQM3KNusmU1_kyKYdkv6lpo,15591
+geney/tcga_utils.py,sha256=uAjejr7F-XqcXS5uANGlsHLOlzMmGo4CTbWhMO0E318,15589
 geney/utils.py,sha256=YOe22gA0Oew9_QEym7ivM9sb7t3wNeHTeiSDBmvOPso,1984
 geney/analyzers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 geney/analyzers/benchmark_clinvar.py,sha256=ZAxvZ-Ue5T6au5mGbk8clfvbAYl13NIY7U92KzL0lXI,5531
@@ -43,7 +44,7 @@ geney/translation_initiation/resources/kozak_pssm.json,sha256=pcd0Olziutq-6H3mFW
 geney/translation_initiation/resources/tis_regressor_model.joblib,sha256=IXb4DUDhJ5rBDKcqMk9zE3ECTZZcdj7Jixz3KpoZ7OA,2592025
 geney/translation_termination/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 geney/translation_termination/tts_utils.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-geney-1.1.10.dist-info/METADATA,sha256=Et_H-jo1c9eYsqEnjk1_gVLRtJYzS9-RXs8RE7Z-u6c,1131
-geney-1.1.10.dist-info/WHEEL,sha256=iYlv5fX357PQyRT2o6tw1bN-YcKFFHKqB_LwHO5wP-g,110
-geney-1.1.10.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
-geney-1.1.10.dist-info/RECORD,,
+geney-1.1.11.dist-info/METADATA,sha256=eKUG3cuHIC37_E6QJg5TyDjBC6NXoine75FZWLxCK6A,1131
+geney-1.1.11.dist-info/WHEEL,sha256=iYlv5fX357PQyRT2o6tw1bN-YcKFFHKqB_LwHO5wP-g,110
+geney-1.1.11.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
+geney-1.1.11.dist-info/RECORD,,