Fix grouping error in cyclic deconvoluter

examples/waters_cyclic_deconvolute.py

@@ -286,7 +286,7 @@ def __init__(self, ms_file,
 
         if storage_type is None:
             storage_type = IonMobilityAware3DThreadedMzMLScanStorageHandler
-
+        self.stream_cls = None
         self.ms_file = ms_file
         self.storage_path = storage_path
         self.sample_name = sample_name
@@ -398,8 +398,7 @@ def cli():
 def feature_deconvolution(input_path, output_path, lockmass_config=None, start_time=0, end_time=None, averagine='glycopeptide',
                           minimum_intensity=10.0, lock_mass_function=3, processes: int = 4,
                           isolation_window_width=0.0, denoise=1.0, signal_averaging=2):
-    """Extract features from each IM-MS cycle followed by deisotoping and charge state deconvolution.
-    """
+    """Extract features from each IM-MS cycle followed by deisotoping and charge state deconvolution."""
     sample_name = os.path.basename(input_path).rsplit(".", 1)[0]
     logging.basicConfig(
         level="INFO", format='%(asctime)s %(message)s', datefmt='%m/%d/%Y %I:%M:%S %p',
@@ -429,10 +428,9 @@ def feature_deconvolution(input_path, output_path, lockmass_config=None, start_t
 
     task = MSESampleConsumer(
         input_path, storage_path=output_path,
-        ms1_peak_picking_args={"error_tolerance": 4e-5,
-                               "minimum_intensity": minimum_intensity / 2, "denoise": denoise},
+        ms1_peak_picking_args={"minimum_intensity": minimum_intensity / 2, "denoise": denoise},
         msn_peak_picking_args={
-            "average_within": signal_averaging, "error_tolerance": 4e-5, "denoise": denoise},
+            "average_within": signal_averaging, "denoise": denoise},
         ms1_deconvolution_args={
             "averagine": averagine,
             "truncate_after": 0.95,
@@ -469,7 +467,8 @@ def feature_deconvolution(input_path, output_path, lockmass_config=None, start_t
 @click.option("-w", "--weight-scaling", type=float, default=1.0,
               help="A weight scaling factor. >1 to incrase the overall abundance of deconvolved peaks")
 def precursor_product_deconvolution(input_path, output_path, edges_per_feature: int=1000, weight_scaling: float=1.0):
-    """Takes a deconvolved LC-IM-MSe run and generate pseudospectra for
+    """
+    Takes a deconvolved LC-IM-MSe run and generate pseudospectra for
     precursor ions using correlated product ion features enclosed in the IM
     and RT dimensions.
     """
@@ -534,7 +533,8 @@ def precursor_product_deconvolution(input_path, output_path, edges_per_feature:
 @click.argument("input_path", type=click.Path())
 @click.argument("output_path", type=click.Path(writable=True))
 def naive_ion_mobility_overlap_pseudospectra(input_path, output_path):
-    """Generate pseudo-spectra from deconvolved cycles where the precursor ion spans
+    """
+    Generate pseudo-spectra from deconvolved cycles where the precursor ion spans
     some or all of the ion dimension of the product ion's mobility dimension and has
     a larger neutral mass. Makes no use of retention time.
     """

src/ms_deisotope/data_source/_vendor/masslynx/loader.py

@@ -4,7 +4,7 @@
 import sys
 
 from collections import defaultdict, OrderedDict
-from typing import Iterator
+from typing import Any, Iterator
 
 import numpy as np
 from ms_deisotope.data_source.metadata.data_transformation import DataProcessingInformation, ProcessingMethod
@@ -40,7 +40,8 @@
 
 
 def is_waters_raw_dir(path):
-    """Detect whether or not the file referenced by ``path``
+    """
+    Detect whether or not the file referenced by ``path``
     is a Waters RAW directory.
 
     Parameters
@@ -66,7 +67,8 @@ def is_waters_raw_dir(path):
 
 
 def infer_reader(path):
-    """If the file referenced by ``path`` is a Waters RAW
+    """
+    If the file referenced by ``path`` is a Waters RAW
     directory, return the callable (:class:`MassLynxRawLoader`) to
     open it, otherwise raise an exception.
 
@@ -431,6 +433,7 @@ def __init__(self, raw_path, lockmass_config=None, default_isolation_width=0.0,
             self.source_file)
         self.lockmass_processor = MassLynxLockMassProcessor.MassLynxLockMassProcessor()
         self.lockmass_processor.SetRawData(self.scan_reader)
+        self.lockmass_config = lockmass_config
         self.configure_lockmass(lockmass_config)
         self.index = []
         self.cycle_index = []
@@ -467,6 +470,9 @@ def configure_lockmass(self, lockmass_config=None):
         if self.lockmass_processor.CanLockMassCorrect():
             self.lockmass_processor.LockMassCorrect()
 
+    def __reduce__(self):
+        return self.__class__, (self.source_file, self.lockmass_config, self.default_isolation_width)
+
     def __repr__(self):
         return "MassLynxRawLoader(%r)" % (self.source_file)
 

src/ms_deisotope/data_source/scan/mobility_frame.py

@@ -195,7 +195,8 @@ def __call__(self, value):
 
 
 class RawDataArrays3D(namedtuple("RawDataArrays3D", ['mz', 'intensity', 'ion_mobility'])):
-    """Represent the m/z, intensity, and ion mobility arrays associated with a raw
+    """
+    Represent the m/z, intensity, and ion mobility arrays associated with a raw
     ion mobility frame of mass spectra.
 
     Thin wrapper around a ``namedtuple``, so this object supports
@@ -233,7 +234,8 @@ def drift_time(self):
         return self.ion_mobility
 
     def has_array(self, array_type):
-        """Check if this array set contains an array of the
+        """
+        Check if this array set contains an array of the
         requested type.
 
         This method uses the semantic lookup mechanism to test
@@ -286,7 +288,8 @@ def __copy__(self):
         return inst
 
     def copy(self):
-        """Make a deep copy of this object.
+        """
+        Make a deep copy of this object.
 
         Returns
         -------
@@ -322,7 +325,8 @@ def __getitem__(self, i):
 
     @classmethod
     def stack(cls, scans, ion_mobility_array_type=None):
-        """Combine multiple :class:`~.Scan` objects or (ion mobility, :class:`~.RawDataArrays`)
+        """
+        Combine multiple :class:`~.Scan` objects or (ion mobility, :class:`~.RawDataArrays`)
         pairs into a single :class:`~.RawDataArrays3D`
 
         Parameters
@@ -367,7 +371,8 @@ def stack(cls, scans, ion_mobility_array_type=None):
 
     @classmethod
     def from_arrays(cls, mz_array, intensity_array, ion_mobility_array, ion_mobility_array_type=None, data_arrays=None):
-        """Build a new :class:`~.RawDataArrays3D` from parallel arrays.
+        """
+        Build a new :class:`~.RawDataArrays3D` from parallel arrays.
 
         This will sort all arrays w.r.t. m/z and ion mobility.
 
@@ -414,7 +419,8 @@ def _slice(self, i, include_ion_mobility=True):
             self.ion_mobility_array_type, data_arrays)
 
     def unstack(self, include_empty=True):
-        """Convert this 3D array into a list of (ion mobility, :class:`~.RawDataArrays`) pairs
+        """
+        Convert this 3D array into a list of (ion mobility, :class:`~.RawDataArrays`) pairs
 
         Parameters
         ----------
@@ -448,7 +454,8 @@ def unstack(self, include_empty=True):
         return acc
 
     def grid(self):
-        """Convert this 3D array into an intensity grid with
+        """
+        Convert this 3D array into an intensity grid with
         m/z along the rows and ion mobility along the columns.
 
         Returns
@@ -469,7 +476,8 @@ def grid(self):
         return mz_axis, im_axis, intensity
 
     def find_mz(self, mz):
-        """Find the nearest index to the query ``mz``
+        """
+        Find the nearest index to the query ``mz``
 
         Parameters
         ----------
@@ -524,7 +532,8 @@ def find_mz(self, mz):
         return 0
 
     def between_mz(self, low, high):
-        """Return a slice of the arrays between ``low`` and ``high`` m/z.
+        """
+        Return a slice of the arrays between ``low`` and ``high`` m/z.
 
         Parameters
         ----------
@@ -622,7 +631,8 @@ def to_deconvoluted_peak_set(self, time_bound=None, mass_upper_bound: float = fl
 
 
 class IonMobilityFrame(FrameBase):
-    """A :class:`IonMobilityFrame` represents an single time point acquisition of
+    """
+    A :class:`IonMobilityFrame` represents an single time point acquisition of
     multiple mass spectra across multiple ion mobility drift time points. Because
     of its drift time by m/z structure, it does not have 1-D peak sets, but pseudo-
     :class:`~.LCMSFeatureMap`-like data structures which conserve the over-time

src/ms_deisotope/feature_map/mobility_frame_processor.py

@@ -82,7 +82,8 @@ def __init__(self, data_source, ms1_peak_picking_args=None,
 
     @property
     def reader(self) -> IonMobilitySourceRandomAccessFrameSource:
-        """The :class:`~.IonMobilitySourceRandomAccessFrameSource` which generates the raw scans that will
+        """
+        The :class:`~.IonMobilitySourceRandomAccessFrameSource` which generates the raw scans that will
         be processed.
 
         Returns

src/ms_deisotope/processor.py

@@ -180,6 +180,8 @@ def _loader_creator(specification, **kwargs):
         raise ValueError("Cannot determine how to get a ScanIterator from %r" % (specification,))
 
 
+# NOTE: This simplification step leads to potentially large deviations from the "unthresholded"
+# deconvolution solution. This procedure is too aggressive for real world data.
 def _simplify_peak_set(peaks, bin_width=5.0):
     bin_edges = np.arange(0, peaks[-1].mz + bin_width, bin_width)
     bins = []
@@ -428,7 +430,8 @@ def _get_envelopes(self, precursor_scan: Scan) -> Optional[List[Tuple[float, flo
         return chosen_envelopes
 
     def _pick_precursor_scan_peaks(self, precursor_scan: Scan, chosen_envelopes: Optional[List[Tuple[float, float]]]=None) -> Union[PeakIndex, PeakSet]:
-        """Pick peaks from the given precursor scan
+        """
+        Pick peaks from the given precursor scan
 
         Parameters
         ----------
@@ -457,7 +460,8 @@ def _pick_precursor_scan_peaks(self, precursor_scan: Scan, chosen_envelopes: Opt
         return prec_peaks
 
     def _average_ms1(self, precursor_scan: Scan) -> Union[PeakIndex, PeakSet]:
-        """Average signal from :attr:`self.ms1_averaging` scans from
+        """
+        Average signal from :attr:`self.ms1_averaging` scans from
         before and after ``precursor_scan`` and pick peaks from the
         averaged arrays.
 
@@ -478,7 +482,8 @@ def _average_ms1(self, precursor_scan: Scan) -> Union[PeakIndex, PeakSet]:
         return prec_peaks
 
     def pick_precursor_scan_peaks(self, precursor_scan: Scan) -> Union[PeakIndex, PeakSet]:
-        """Picks peaks for the given ``precursor_scan`` using the
+        """
+        Picks peaks for the given ``precursor_scan`` using the
         appropriate strategy.
 
         If :attr:`ms1_averaging` > 0, then the signal averaging strategy
@@ -487,6 +492,7 @@ def pick_precursor_scan_peaks(self, precursor_scan: Scan) -> Union[PeakIndex, Pe
         Parameters
         ----------
         precursor_scan: Scan
+            The scan to pick peaks from
 
         Returns
         -------
@@ -498,6 +504,8 @@ def pick_precursor_scan_peaks(self, precursor_scan: Scan) -> Union[PeakIndex, Pe
         else:
             prec_peaks = self._pick_precursor_scan_peaks(precursor_scan)
         n_peaks = len(prec_peaks)
+        # NOTE: This simplification step leads to potentially large deviations from the "unthresholded"
+        # deconvolution solution. This procedure is too aggressive for real world data.
         if n_peaks > self.too_many_peaks_threshold:
             self.log("%d peaks found for %r, applying local intensity threshold." % (n_peaks, precursor_scan))
             prec_peaks = _simplify_peak_set(prec_peaks)