From 7205db5c099c12d752cbf7183c995ac9b985b330 Mon Sep 17 00:00:00 2001
From: Tim O'Donnell <timodonnell@gmail.com>
Date: Tue, 19 Nov 2019 18:05:11 -0500
Subject: [PATCH] fixes
---
mhcflurry/allele_encoding.py | 14 +++
mhcflurry/class1_ligandome_predictor.py | 124 ++++++++++++++++++++----
mhcflurry/custom_loss.py | 26 ++---
mhcflurry/random_negative_peptides.py | 6 ++
test/test_class1_ligandome_predictor.py | 5 +-
5 files changed, 134 insertions(+), 41 deletions(-)
diff --git a/mhcflurry/allele_encoding.py b/mhcflurry/allele_encoding.py
index 9974baca..273f8fcb 100644
--- a/mhcflurry/allele_encoding.py
+++ b/mhcflurry/allele_encoding.py
@@ -208,3 +208,17 @@ class MultipleAlleleEncoding(object):
def fixed_length_vector_encoded_sequences(self, encoding_name):
raise NotImplementedError()
+
+ def shuffle_in_place(self, shuffle_permutation=None):
+ flattened_alleles = self.allele_encoding.alleles.values
+ reshaped_alleles = numpy.reshape(
+ flattened_alleles, (-1, self.max_alleles_per_experiment))
+
+ if shuffle_permutation is None:
+ shuffle_permutation = numpy.random.permutation(
+ len(reshaped_alleles))
+
+ self.allele_encoding = AlleleEncoding(
+ alleles=reshaped_alleles[shuffle_permutation].flatten(),
+ borrow_from=self.allele_encoding
+ )
\ No newline at end of file
diff --git a/mhcflurry/class1_ligandome_predictor.py b/mhcflurry/class1_ligandome_predictor.py
index f1d34505..32c2c1df 100644
--- a/mhcflurry/class1_ligandome_predictor.py
+++ b/mhcflurry/class1_ligandome_predictor.py
@@ -12,11 +12,12 @@ from .class1_neural_network import Class1NeuralNetwork, DEFAULT_PREDICT_BATCH_SI
from .encodable_sequences import EncodableSequences
from .regression_target import from_ic50, to_ic50
from .random_negative_peptides import RandomNegativePeptides
+from .allele_encoding import MultipleAlleleEncoding
from .custom_loss import (
MSEWithInequalities,
- MSEWithInequalitiesAndMultipleOutputs,
MultiallelicMassSpecLoss)
+
class Class1LigandomePredictor(object):
network_hyperparameter_defaults = HyperparameterDefaults(
allele_amino_acid_encoding="BLOSUM62",
@@ -37,7 +38,8 @@ class Class1LigandomePredictor(object):
max_epochs=500,
validation_split=0.1,
early_stopping=True,
- minibatch_size=128,).extend(
+ minibatch_size=128,
+ random_negative_affinity_min=20000.0,).extend(
RandomNegativePeptides.hyperparameter_defaults
)
"""
@@ -273,26 +275,47 @@ class Class1LigandomePredictor(object):
# layer.trainable = False
# import ipdb ; ipdb.set_trace()
- peptides = EncodableSequences.create(peptides)
- peptide_encoding = self.peptides_to_network_input(peptides)
-
- # Optional optimization
- allele_encoding = allele_encoding.compact()
-
- (allele_encoding_input, allele_representations) = (
- self.allele_encoding_to_network_input(allele_encoding))
+ encodable_peptides = EncodableSequences.create(peptides)
+ if labels is not None:
+ labels = numpy.array(labels, copy=False)
+ if affinities_mask is not None:
+ affinities_mask = numpy.array(affinities_mask, copy=False)
+ if inequalities is not None:
+ inequalities = numpy.array(inequalities, copy=False)
+
+ random_negatives_planner = RandomNegativePeptides(
+ **RandomNegativePeptides.hyperparameter_defaults.subselect(
+ self.hyperparameters))
+ random_negatives_planner.plan(
+ peptides=encodable_peptides.sequences[affinities_mask],
+ affinities=(labels[affinities_mask]),
+ alleles=numpy.reshape(
+ allele_encoding.allele_encoding.alleles.values,
+ (-1, allele_encoding.max_alleles_per_experiment))[
+ affinities_mask, 0
+ ].flatten(),
+ inequalities=inequalities[affinities_mask])
+
+ peptide_input = self.peptides_to_network_input(encodable_peptides)
# Shuffle
if shuffle_permutation is None:
shuffle_permutation = numpy.random.permutation(len(labels))
- peptide_encoding = peptide_encoding[shuffle_permutation]
- allele_encoding_input = allele_encoding_input[shuffle_permutation]
+ peptide_input = peptide_input[shuffle_permutation]
+ allele_encoding.shuffle_in_place(shuffle_permutation)
labels = labels[shuffle_permutation]
inequalities = inequalities[shuffle_permutation]
affinities_mask = affinities_mask[shuffle_permutation]
+ del peptides
+ del encodable_peptides
- x_dict = {
- 'peptide': peptide_encoding,
+ # Optional optimization
+ allele_encoding = allele_encoding.compact()
+ (allele_encoding_input, allele_representations) = (
+ self.allele_encoding_to_network_input(allele_encoding))
+
+ x_dict_without_random_negatives = {
+ 'peptide': peptide_input,
'allele': allele_encoding_input,
}
@@ -300,6 +323,18 @@ class Class1LigandomePredictor(object):
if affinities_mask is not None:
y1[affinities_mask] = from_ic50(labels[affinities_mask])
+ random_negatives_allele_encoding = None
+ if allele_encoding is not None:
+ random_negative_alleles = random_negatives_planner.get_alleles()
+ random_negatives_allele_encoding = MultipleAlleleEncoding(
+ experiment_names=random_negative_alleles,
+ experiment_to_allele_list=dict(
+ (a, [a]) for a in random_negative_alleles),
+ max_alleles_per_experiment=(
+ allele_encoding.max_alleles_per_experiment),
+ borrow_from=allele_encoding.allele_encoding)
+ num_random_negatives = random_negatives_planner.get_total_count()
+
if inequalities is not None:
# Reverse inequalities because from_ic50() flips the direction
# (i.e. lower affinity results in higher y values).
@@ -313,16 +348,35 @@ class Class1LigandomePredictor(object):
adjusted_inequalities[~affinities_mask] = ">"
+ # Note: we are using "<" here not ">" because the inequalities are
+ # now in target-space (0-1) not affinity-space.
+ adjusted_inequalities_with_random_negative = numpy.concatenate([
+ numpy.tile("<", num_random_negatives),
+ adjusted_inequalities
+ ])
+ random_negative_ic50 = self.hyperparameters[
+ 'random_negative_affinity_min'
+ ]
+ y1_with_random_negatives = numpy.concatenate([
+ numpy.tile(
+ from_ic50(random_negative_ic50), num_random_negatives),
+ y1,
+ ])
+
affinities_loss = MSEWithInequalities()
encoded_y1 = affinities_loss.encode_y(
- y1, inequalities=adjusted_inequalities)
+ y1_with_random_negatives,
+ inequalities=adjusted_inequalities_with_random_negative)
mms_loss = MultiallelicMassSpecLoss(
delta=self.hyperparameters['loss_delta'])
-
y2 = labels.copy()
y2[affinities_mask] = -1
- encoded_y2 = mms_loss.encode_y(y2)
+ y2_with_random_negatives = numpy.concatenate([
+ numpy.tile(0.0, num_random_negatives),
+ y2,
+ ])
+ encoded_y2 = mms_loss.encode_y(y2_with_random_negatives)
fit_info = collections.defaultdict(list)
@@ -344,13 +398,45 @@ class Class1LigandomePredictor(object):
min_val_loss = None
last_progress_print = 0
start = time.time()
+ x_dict_with_random_negatives = {}
for i in range(self.hyperparameters['max_epochs']):
epoch_start = time.time()
+ random_negative_peptides = EncodableSequences.create(
+ random_negatives_planner.get_peptides())
+ random_negative_peptides_encoding = (
+ self.peptides_to_network_input(random_negative_peptides))
+
+ if not x_dict_with_random_negatives:
+ if len(random_negative_peptides) > 0:
+ x_dict_with_random_negatives[
+ "peptide"
+ ] = numpy.concatenate([
+ random_negative_peptides_encoding,
+ x_dict_without_random_negatives['peptide'],
+ ])
+ x_dict_with_random_negatives[
+ 'allele'
+ ] = numpy.concatenate([
+ self.allele_encoding_to_network_input(
+ random_negatives_allele_encoding)[0],
+ x_dict_without_random_negatives['allele']
+ ])
+ else:
+ x_dict_with_random_negatives = (
+ x_dict_without_random_negatives)
+ else:
+ # Update x_dict_with_random_negatives in place.
+ # This is more memory efficient than recreating it as above.
+ if len(random_negative_peptides) > 0:
+ x_dict_with_random_negatives[
+ "peptide"
+ ][:num_random_negatives] = random_negative_peptides_encoding
+
# TODO: need to use fit_generator to keep each minibatch corresponding
# to a single experiment
fit_history = self.network.fit(
- x_dict,
+ x_dict_with_random_negatives,
[encoded_y1, encoded_y2],
shuffle=True,
batch_size=self.hyperparameters['minibatch_size'],
@@ -410,7 +496,7 @@ class Class1LigandomePredictor(object):
progress_callback()
fit_info["time"] = time.time() - start
- fit_info["num_points"] = len(peptides)
+ fit_info["num_points"] = len(encodable_peptides)
self.fit_info.append(dict(fit_info))
def predict(
diff --git a/mhcflurry/custom_loss.py b/mhcflurry/custom_loss.py
index 84a83459..2434bfaf 100644
--- a/mhcflurry/custom_loss.py
+++ b/mhcflurry/custom_loss.py
@@ -235,12 +235,6 @@ class MSEWithInequalitiesAndMultipleOutputs(Loss):
class MultiallelicMassSpecLoss(Loss):
"""
- Affiniities are mapped according to MSEWithInequalities, then by:
- x -> x + 10.
-
- Mass spec hit vs. decoy are kept at [0, 1].
-
-
"""
name = "multiallelic_mass_spec_loss"
supports_inequalities = True
@@ -251,20 +245,12 @@ class MultiallelicMassSpecLoss(Loss):
@staticmethod
def encode_y(y, affinities_mask=None, inequalities=None):
- y = array(y, dtype="float32")
- if isnan(y).any():
- raise ValueError("y contains NaN", y)
- if (y > 1.0).any():
- raise ValueError("y contains values > 1.0", y)
- if (y < 0.0).any():
- raise ValueError("y contains values < 0.0", y)
-
- encoded = y.copy()
- if affinities_mask is not None:
- encoded[affinities_mask] = MSEWithInequalities.encode_y(
- encoded[affinities_mask], inequalities=inequalities) + 10.0
-
- return encoded
+ encoded = pandas.Series(y, dtype="float32", copy=True)
+ assert all(item in (-1.0, 1.0, 0.0) for item in encoded), set(y)
+ print(
+ "MultiallelicMassSpecLoss: y-value counts:\n",
+ encoded.value_counts())
+ return encoded.values
def loss(self, y_true, y_pred):
import tensorflow as tf
diff --git a/mhcflurry/random_negative_peptides.py b/mhcflurry/random_negative_peptides.py
index b29adc9c..978a93e3 100644
--- a/mhcflurry/random_negative_peptides.py
+++ b/mhcflurry/random_negative_peptides.py
@@ -70,6 +70,12 @@ class RandomNegativePeptides(object):
pandas.DataFrame indicating number of random negatives for each length
and allele.
"""
+ numpy.testing.assert_equal(len(peptides), len(affinities))
+ if alleles is not None:
+ numpy.testing.assert_equal(len(peptides), len(alleles))
+ if inequalities is not None:
+ numpy.testing.assert_equal(len(peptides), len(inequalities))
+
peptides = pandas.Series(peptides, copy=False)
peptide_lengths = peptides.str.len()
diff --git a/test/test_class1_ligandome_predictor.py b/test/test_class1_ligandome_predictor.py
index f081d09b..41ff42ff 100644
--- a/test/test_class1_ligandome_predictor.py
+++ b/test/test_class1_ligandome_predictor.py
@@ -282,10 +282,11 @@ def test_real_data_multiallelic_refinement(max_epochs=10):
ligandome_predictor = Class1LigandomePredictor(
pan_predictor,
max_ensemble_size=1,
- max_epochs=500,
+ max_epochs=50,
learning_rate=0.0001,
patience=5,
- min_delta=0.0)
+ min_delta=0.0,
+ random_negative_rate=1.0)
pre_predictions = from_ic50(
ligandome_predictor.predict(
--
GitLab