pymaid.adjacency_matrix

pymaid.adjacency_matrix(sources, targets=None, source_grp={}, target_grp={}, fractions=False, syn_threshold=None, syn_cutoff=None, use_connectors=False, volume_filter=None, remote_instance=None)

Generate adjacency matrix between source and target neurons.

Directional: sources = rows, targets = columns.

Parameters:

• sources –

  Source neurons as single or list of either:

  1. skeleton IDs (int or str)

  2. neuron name (str, exact match)

  3. annotation: e.g. ‘annotation:PN right’

  4. CatmaidNeuron or CatmaidNeuronList object

• targets –

  Optional. Target neurons as single or list of either:

  1. skeleton IDs (int or str)

  2. neuron name (str, exact match)

  3. annotation: e.g. ``’annotation:PN right’`

  4. CatmaidNeuron or CatmaidNeuronList object

  If not provided, source neurons = target neurons.

• fractions (bool, optional) – If True, will return connectivity as fraction of total number of postsynaptic links to target neuron.

• syn_cutoff (int, optional) – If set, will cut off connections ABOVE given value.

• syn_threshold (int, optional) – If set, will ignore connections with LESS synapses.

• source_grp (dict, optional) –

  Use to collapse sources into groups. Can be either:

  1. {group1: [neuron1, neuron2, ... ], ..}

  2. {neuron1: group1, neuron2 : group2, ..}

  syn_cutoff and syn_threshold are applied BEFORE grouping!

• target_grp (dict, optional) – See source_grp for possible formats.

• use_connectors (bool, optional) – If True AND s or t are CatmaidNeuron/List, restrict adjacency matrix to their connectors. Use if e.g. you are using pruned neurons.

• volume_filter (Volume | list of Volumes, optional) – Volume(s) to restrict connections to. Can be a pymaid.Volume, the name of a CATMAID volume or a list thereof.

• remote_instance (CatmaidInstance, optional) – If not passed, will try using globally defined.

Returns:

matrix

Return type:

pandas.Dataframe

See also

group_matrix()

More fine-grained control over matrix grouping.

adjacency_from_connectors()

Use this function if you are working with multiple fragments per neuron.

Examples

Generate and plot a adjacency matrix:

>>> import seaborn as sns
>>> import matplotlib.pyplot as plt
>>> neurons = pymaid.get_neurons('annotation:test')
>>> mat = pymaid.adjacency_matrix(neurons)
>>> g = sns.heatmap(adj_mat, square=True)
>>> g.set_yticklabels(g.get_yticklabels(), rotation=0, fontsize=7)
>>> g.set_xticklabels(g.get_xticklabels(), rotation=90, fontsize=7)
>>> plt.show()

Cut neurons into axon dendrites and compare their connectivity:

>>> # Get a set of neurons
>>> nl = pymaid.get_neurons('annnotation:type_16_candidates')
>>> # Split into axon dendrite by using a tag
>>> nl.reroot(nl.soma)
>>> nl_axon = nl.prune_proximal_to('axon', inplace=False)
>>> nl_dend = nl.prune_distal_to('axon', inplace=False)
>>> # Get a list of the downstream partners
>>> cn_table = pymaid.get_partners(nl)
>>> ds_partners = cn_table[cn_table.relation == 'downstream']
>>> # Take the top 10 downstream partners
>>> top_ds = ds_partners.iloc[:10].skeleton_id.values
>>> # Generate separate adjacency matrices for axon and dendrites
>>> adj_axon = pymaid.adjacency_matrix(nl_axon, top_ds,
...                                    use_connectors=True)
>>> adj_dend = pymaid.adjacency_matrix(nl_dend, top_ds,
...                                    use_connectors=True)
>>> # Rename rows and merge dataframes
>>> adj_axon.index += '_axon'
>>> adj_dend.index += '_dendrite'
>>> adj_merged = pd.concat([adj_axon, adj_dend], axis=0)
>>> # Plot heatmap using seaborn
>>> ax = sns.heatmap(adj_merged)
>>> plt.show()

Restrict adjacency matrix to a given volume:

>>> neurons = pymaid.get_neurons('annotation:glomerulus DA1')
>>> lh = pymaid.get_volume('LH_R')
>>> adj = pymaid.adjacency_matrix(neurons, volume_filter=lh)

Get adjacency matrix with fraction of inputs instead of total synapse count:

>>> neurons = pymaid.get_neurons('annotation:glomerulus DA1')
>>> adj = pymaid.adjacency_matrix(neurons, fractions=True)