Merge pull request #372 from AllenInstitute/update/workshop-2024

Update/workshop 2024

Author: kaeldai

bmtk/analyzer/ecp.py

@@ -2,10 +2,12 @@
 import h5py
 import matplotlib.pyplot as plt
 import numpy as np
+from decimal import Decimal
 
 from bmtk.utils.sonata.config import SonataConfig
 from bmtk.simulator.utils import simulation_reports
-# from mpl_toolkits.axes_grid1.anchored_artists import AnchoredSizeBar
+from mpl_toolkits.axes_grid1.anchored_artists import AnchoredSizeBar
+import matplotlib.font_manager as fm
 
 
 def _get_ecp_path(ecp_path=None, config=None, report_name=None):
@@ -55,30 +57,54 @@ def plot_ecp(config_file=None, report_name=None, ecp_path=None, title=None, show
     channels = ecp_h5['/ecp/channel_id'][()]
     fig, axes = plt.subplots(len(channels), 1)
     fig.text(0.04, 0.5, 'channel id', va='center', rotation='vertical')
+    v_min, v_max = ecp_h5['/ecp/data'][()].min(), ecp_h5['/ecp/data'][()].max()
+    # print(v_max - v_min)
+    # exit()
+    
     for idx, channel in enumerate(channels):
         data = ecp_h5['/ecp/data'][:, idx]
+        # print(channel, np.min(data), np.max(data))
         axes[idx].plot(time_traces, data)
         axes[idx].spines["top"].set_visible(False)
         axes[idx].spines["right"].set_visible(False)
         axes[idx].set_yticks([])
         axes[idx].set_ylabel(channel)
+        axes[idx].set_ylim([v_min, v_max])
 
         if idx+1 != len(channels):
             axes[idx].spines["bottom"].set_visible(False)
             axes[idx].set_xticks([])
         else:
             axes[idx].set_xlabel('timestamps (ms)')
-            # scalebar = AnchoredSizeBar(axes[idx].transData,
-            #                            2.0, '1 mV', 1,
-            #                            pad=0,
-            #                            borderpad=0,
-            #                            # color='b',
-            #                            frameon=True,
-            #                            # size_vertical=1.001,
-            #                            # fontproperties=fontprops
-            #                            )
-            #
-            # axes[idx].add_artist(scalebar)
+        
+        
+        if idx == 0:
+            scale_bar_size = (v_max-v_min)/2.0
+            scale_bar_label = f'{scale_bar_size:.2E}'
+            # print(scale_bar_label)
+            # exit()
+            fontprops = fm.FontProperties(size='x-small')
+
+            scalebar = AnchoredSizeBar(
+                axes[idx].transData,
+                size=scale_bar_size, 
+                label=scale_bar_label, 
+                loc='upper right',
+                pad=0.1,
+                borderpad=0.5,
+                sep=5,
+                # color='b',
+                frameon=False,
+                size_vertical=scale_bar_size,
+                # size_vertical=1.001,
+                fontproperties=fontprops
+            )            
+            axes[idx].add_artist(scalebar)
+
+            # label = scalebar.txt_label
+            # label.set_rotation(270.0)
+            # label.set_verticalalignment('bottom')
+            # label.set_horizontalalignment('left') 
 
     if title:
         fig.set_title(title)

bmtk/simulator/bionet/__init__.py

@@ -20,7 +20,8 @@
 # WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #
-from bmtk.simulator.bionet.pyfunction_cache import synapse_model, synaptic_weight, cell_model, add_weight_function, model_processing
+from bmtk.simulator.bionet.pyfunction_cache import synapse_model, synaptic_weight, cell_model, add_weight_function, model_processing, \
+    spikes_generator
 from bmtk.simulator.bionet.config import Config
 from bmtk.simulator.bionet.bionetwork import BioNetwork
 from bmtk.simulator.bionet.biosimulator import BioSimulator

bmtk/simulator/bionet/biocell.py

@@ -27,6 +27,7 @@
 from bmtk.simulator.bionet.morphology import Morphology
 import six
 
+import neuron
 from neuron import h
 
 pc = h.ParallelContext()    # object to access MPI methods
@@ -74,9 +75,6 @@ class BioCell(Cell):
     def __init__(self, node, population_name, bionetwork):
         super(BioCell, self).__init__(node=node, population_name=population_name, network=bionetwork)
 
-        # Set up netcon object that can be used to detect and communicate cell spikes.
-        self.set_spike_detector(bionetwork.spike_threshold)
-
         # Determine number of segments and store a list of all sections.
         self._secs = []
         self._secs_by_id = []
@@ -105,6 +103,10 @@ def __init__(self, node, population_name, bionetwork):
         self._seg_coords = None
         self.build_morphology()
 
+        # Set up netcon object that can be used to detect and communicate cell spikes.
+        self.set_spike_detector(bionetwork.spike_threshold)
+
+
     def build_morphology(self):
         morph_base = Morphology.load(hobj=self.hobj, morphology_file=self.morphology_file, cache_seg_props=True)
 
@@ -126,6 +128,10 @@ def morphology(self):
         """The actual Morphology object instanstiation"""
         return self._morphology
 
+    @property
+    def soma(self):
+        return self.morphology.soma
+
     @property
     def seg_coords(self):
         """Coordinates for segments/sections of the morphology, need to make public for ecp, xstim, and other
@@ -144,7 +150,7 @@ def seg_coords(self):
         return self.morphology.seg_coords
 
     def set_spike_detector(self, spike_threshold):
-        nc = h.NetCon(self.hobj.soma[0](0.5)._ref_v, None, sec=self.hobj.soma[0])  # attach spike detector to cell
+        nc = h.NetCon(self.soma[0](0.5)._ref_v, None, sec=self.soma[0])
         nc.threshold = spike_threshold
         pc.cell(self.gid, nc)  # associate gid with spike detector
 
@@ -437,18 +443,41 @@ def __init__(self, node, population_name, bionetwork):
         self._vecstim = h.VecStim()
         self._vecstim.play(self._spike_trains)
 
-        self._precell_filter = bionetwork.spont_syns_filter
+        self._precell_filter = bionetwork.spont_syns_filter_pre
+        self._postcell_filter = bionetwork.spont_syns_filter_post
         assert(isinstance(self._precell_filter, dict))
 
-    def _matches_filter(self, src_node):
+    def _matches_filter(self, src_node, trg_node=None):
         """Check to see if the presynaptic cell matches the criteria specified"""
         for k, v in self._precell_filter.items():
+            # Some key may not show up as node_variable
+            if k == 'population' and k not in src_node:
+                key_val = src_node.population_name
+            else:
+                key_val = src_node[k]
+
+            if isinstance(v, (list, tuple)):
+                if key_val not in v:
+                    return False
+            else:
+                if key_val != v:
+                    return False
+        
+        trg_node = trg_node or self
+        for k, v in self._postcell_filter.items():
+            # Some key may not show up as node_variable
+            if k == 'population' and k not in trg_node:
+                key_val = trg_node._node.population_name
+            else:
+                key_val = trg_node[k]
+
             if isinstance(v, (list, tuple)):
-                if src_node[k] not in v:
+                if key_val not in v:
                     return False
             else:
-                if src_node[k] != v:
+                if key_val != v:
                     return False
+        
         return True
 
     def _set_connections(self, edge_prop, src_node, syn_weight, stim=None):

bmtk/simulator/bionet/bionetwork.py

@@ -77,7 +77,8 @@ def __init__(self):
         self._gid_pool = GidPool()
 
         self.has_spont_syns = False
-        self.spont_syns_filter = None
+        self.spont_syns_filter_pre = None
+        self.spont_syns_filter_post = None
         self.spont_syns_times = None
 
     @property
@@ -88,7 +89,7 @@ def gid_pool(self):
     def py_function_caches(self):
         return nrn
 
-    def set_spont_syn_activity(self, precell_filter, timestamps):
+    def set_spont_syn_activity(self, precell_filter, postcell_filter, timestamps):
         self._model_type_map = {
             'biophysical': BioCellSpontSyn,
             'point_process': PointProcessCellSpontSyns,
@@ -98,7 +99,8 @@ def set_spont_syn_activity(self, precell_filter, timestamps):
         }
 
         self.has_spont_syns = True
-        self.spont_syns_filter = precell_filter
+        self.spont_syns_filter_pre = precell_filter
+        self.spont_syns_filter_post = postcell_filter
         self.spont_syns_times = timestamps
 
     def get_node_id(self, population, node_id):
@@ -134,12 +136,12 @@ def add_nodes(self, node_population):
         self._gid_pool.add_pool(node_population.name, node_population.n_nodes())
         super(BioNetwork, self).add_nodes(node_population)
 
-    def get_virtual_cells(self, population, node_id, spike_trains):
+    def get_virtual_cells(self, population, node_id, spike_trains, spikes_generator=None, sim=None):
         if node_id in self._virtual_nodes[population]:
             return self._virtual_nodes[population][node_id]
         else:
             node = self.get_node_id(population, node_id)
-            virt_cell = VirtualCell(node, population, spike_trains)
+            virt_cell = VirtualCell(node, population, spike_trains, spikes_generator, sim)
             self._virtual_nodes[population][node_id] = virt_cell
             return virt_cell
 
@@ -151,7 +153,7 @@ def get_disconnected_cell(self, population, node_id, spike_trains):
             virt_cell = self._disconnected_source_cells[population][node_id]
         else:
             node = self.get_node_id(population, node_id)
-            virt_cell = VirtualCell(node, population, spike_trains)
+            virt_cell = VirtualCell(node, population, spike_trains, self)
             self._disconnected_source_cells[population][node_id] = virt_cell
 
         return virt_cell
@@ -369,7 +371,7 @@ def find_edges(self, source_nodes=None, target_nodes=None):
 
         return selected_edges
 
-    def add_spike_trains(self, spike_trains, node_set):
+    def add_spike_trains(self, spike_trains, node_set, spikes_generator=None, sim=None):
         self._init_connections()
 
         src_nodes = [node_pop for node_pop in self.node_populations if node_pop.name in node_set.population_names()]
@@ -379,7 +381,7 @@ def add_spike_trains(self, spike_trains, node_set):
                 if edge_pop.virtual_connections:
                     for trg_nid, trg_cell in self._rank_node_ids[edge_pop.target_nodes].items():
                         for edge in edge_pop.get_target(trg_nid):
-                            src_cell = self.get_virtual_cells(source_population, edge.source_node_id, spike_trains)
+                            src_cell = self.get_virtual_cells(source_population, edge.source_node_id, spike_trains, spikes_generator, sim)
                             trg_cell.set_syn_connection(edge, src_cell, src_cell)
 
                 elif edge_pop.mixed_connections:

bmtk/simulator/bionet/biosimulator.py

@@ -326,6 +326,7 @@ def from_config(cls, config, network, set_recordings=True):
             if sim_input.input_type == 'syn_activity':
                 network.set_spont_syn_activity(
                     precell_filter=sim_input.params['precell_filter'],
+                    postcell_filter=sim_input.params.get('postcell_filter', {}),
                     timestamps=sim_input.params['timestamps']
                 )
 
@@ -346,13 +347,30 @@ def from_config(cls, config, network, set_recordings=True):
 
         # TODO: Need to create a gid selector
         for sim_input in inputs.from_config(config):
-            if sim_input.input_type == 'spikes' and sim_input.module in ['nwb', 'csv', 'sonata']:
+            if sim_input.input_type == 'spikes' and sim_input.module in ['nwb', 'csv', 'sonata', 'h5']:
                 io.log_info('Building virtual cell stimulations for {}'.format(sim_input.name))
                 path = sim_input.params['input_file']
                 spikes = SpikeTrains.load(path=path, file_type=sim_input.module, **sim_input.params)
                 # node_set_opts = sim_input.params.get('node_set', 'all')
                 node_set = network.get_node_set(sim_input.node_set)
-                network.add_spike_trains(spikes, node_set)
+                network.add_spike_trains(
+                    spike_trains=spikes, 
+                    node_set=node_set,
+                    spikes_generator=None,
+                    sim=sim
+                )
+
+            elif sim_input.input_type == 'spikes' and sim_input.module == 'function':
+                io.log_info('Building virtual cell stimulations for {}'.format(sim_input.name))
+                # path = sim_input.params.get['input_file']
+                spikes_generator = sim_input.params['spikes_function']
+                node_set = network.get_node_set(sim_input.node_set)
+                network.add_spike_trains(
+                    spike_trains=None, 
+                    node_set=node_set, 
+                    spikes_generator=spikes_generator,
+                    sim=sim
+                )
 
             elif sim_input.module == 'IClamp':
                 sim.add_mod(mods.IClampMod(input_type=sim_input.input_type, **sim_input.params))

bmtk/simulator/bionet/cell.py

@@ -106,6 +106,12 @@ def get_connection_info(self):
 
     def set_syn_connections(self, edge_prop, src_node, stim=None):
         raise NotImplementedError
+    
+    def get_section(self, sec_name, sec_index):
+        raise NotImplementedError
 
+    def __contains__(self, node_prop):
+        return node_prop in self._node
+    
     def __getitem__(self, node_prop):
         return self._node[node_prop]

bmtk/simulator/bionet/config.py

@@ -39,7 +39,12 @@ def create_output_dir(self):
         io.setup_output_dir(self.output_dir, self.log_file)
 
     def load_nrn_modules(self):
-        nrn.load_neuron_modules(self.mechanisms_dir, self.templates_dir)
+        nrn.load_neuron_modules(
+            mechanisms_dir=self.mechanisms_dir, 
+            templates_dir=self.templates_dir, 
+            default_templates=self.use_default_templates,
+            use_old_import3d=self.use_old_import3d
+        )
 
     def build_env(self):
         self.io = io
@@ -52,3 +57,8 @@ def build_env(self):
 
         pc.barrier()
         self.load_nrn_modules()
+
+    def _set_class_props(self):
+        super(Config, self)._set_class_props()
+        self.use_old_import3d = self.run.get('use_old_import3d', False)
+        self.use_default_templates = self.run.get('use_old_import3d', True)

bmtk/simulator/bionet/default_setters/cell_models.py

@@ -23,6 +23,7 @@
 import os
 import numpy as np
 from neuron import h
+import inspect
 try:
     from sklearn.decomposition import PCA
 except Exception as e:
@@ -41,19 +42,46 @@
 """
 
 def loadHOC(cell, template_name, dynamics_params):
-    # Get template to instantiate
-    template_call = getattr(h, template_name)
-    if dynamics_params is not None and 'params' in dynamics_params:
-        template_params = dynamics_params['params']
-        if isinstance(template_params, list):
-            # pass in a list of parameters
-            hobj = template_call(*template_params)
+    """A Generic function for creating a cell object from a NEURON HOC Template (eg. a *.hoc file with 
+    `begintemplate template_name` in header). It essentially tries to guess the correct parameters that need to be
+    called so may not work the majority of the times and require to be overloaded.
+
+    :param cell: A SONATA node object, can be used as a dict to get individual properties of current cell.
+    :param template_name: name of HOCTemplate as stored in "model_template" attribute (hoc:<template_name>).
+    :param dynamics_params: Dictionary containing contents of cell['dynamics_params'] as loaded from a json file or hdf5. 
+        If cell does not have "dynamics_params" attributes then will be set to None.
+    """
+    try:
+        # Get template to instantiate
+        template_call = getattr(h, template_name)
+    except AttributeError as ae:
+        io.log_error(
+            f'loadHOC was unable to load in Neuron HOC Template "{template_name}, ' 
+            'Make sure appropiate .hoc file is stored in templates_dir.'
+        )
+        raise ae
+
+    try:
+        if dynamics_params is not None and 'params' in dynamics_params:
+            template_params = dynamics_params['params']
+            if isinstance(template_params, list):
+                # pass in a list of parameters
+                hobj = template_call(*template_params)
+            else:
+                # only a single parameter
+                hobj = template_call(template_params)
+        elif cell.morphology_file is not None:
+            # instantiate template with no parameters
+            hobj = template_call(cell.morphology_file)
         else:
-            # only a single parameter
-            hobj = template_call(template_params)
-    else:
-        # instantiate template with no parameters
-        hobj = template_call()
+            hobj = template_call()
+    except RuntimeError as rte:
+        io.log_error(
+            f'bmtk.simualtor.bionet.default_setters.cell_models.loadHOC function failed to load HOC template "{template_call}". '
+            'If Hoc Templates requires special call to be initialized consider using `bmtk.simulator.bionet.add_cell_model()` '
+            'to overwrite this function.'
+        )
+        raise rte
 
     # TODO: All "all" section if it doesn't exist
     # hobj.all = h.SectionList()