DataJoint U24 - Workflow Miniscope¶

Setup¶

First, let's change directories to find the dj_local_conf file.

In [1]:

import os
import numpy as np
import datajoint as dj

if os.path.basename(os.getcwd()) == "notebooks":
    os.chdir("..")
dj.config["custom"]["database.prefix"] = "u24_mini_"

import os import numpy as np import datajoint as dj if os.path.basename(os.getcwd()) == "notebooks": os.chdir("..") dj.config["custom"]["database.prefix"] = "u24_mini_"

Next, we populate the python namespace with the required schemas

In [2]:

from workflow_miniscope.pipeline import miniscope, QualityMetricFigs

from workflow_miniscope.pipeline import miniscope, QualityMetricFigs

[2023-01-10 11:28:51,513][WARNING]: lab.Project and related tables will be removed in a future version of Element Lab. Please use the project schema.
[2023-01-10 11:28:56,262][INFO]: Connecting cbroz@dss-db.datajoint.io:3306
[2023-01-10 11:28:56,711][INFO]: Connected cbroz@dss-db.datajoint.io:3306

Visualizations¶

CaImAn produces a few metrics for estimates. These can be visualized as follows.

In [3]:

key = miniscope.Curation.fetch("KEY", limit=1)[0]
qm = QualityMetricFigs(miniscope, key, dark_mode=True)
print("Available plots:", qm.plot_list)
fig = qm.get_single_fig("r_values")
fig.show(
    "png"
)  # .show('png') is optional. Here, it is used to render the image within a notebook that is embedded in a browser.qm.get_grid()

key = miniscope.Curation.fetch("KEY", limit=1)[0] qm = QualityMetricFigs(miniscope, key, dark_mode=True) print("Available plots:", qm.plot_list) fig = qm.get_single_fig("r_values") fig.show( "png" ) # .show('png') is optional. Here, it is used to render the image within a notebook that is embedded in a browser.qm.get_grid()

Available plots: ['r_values', 'SNR', 'cnn_preds']

A number of metrics are available from CaImAn estimates.

In [4]:

import inspect

[
    i[0]
    for i in inspect.getmembers(qm.estimates)  # see all available class attributes
    if not i[0].startswith("_") and not inspect.ismethod(i[1])  # filter out functions
]

import inspect [ i[0] for i in inspect.getmembers(qm.estimates) # see all available class attributes if not i[0].startswith("_") and not inspect.ismethod(i[1]) # filter out functions ]

Out[4]:

['A',
 'A_thr',
 'Ab',
 'Ab_dense',
 'AtA',
 'AtY_buf',
 'C',
 'CC',
 'CY',
 'C_on',
 'Cf',
 'Cn',
 'F_dff',
 'OASISinstances',
 'R',
 'S',
 'SNR_comp',
 'W',
 'YrA',
 'Yr_buf',
 'b',
 'b0',
 'bl',
 'c1',
 'center',
 'cnn_preds',
 'coordinates',
 'dims',
 'discarded_components',
 'ecc',
 'f',
 'g',
 'groups',
 'idx_components',
 'idx_components_bad',
 'ind_new',
 'lam',
 'merged_ROIs',
 'mn',
 'neurons_sn',
 'noisyC',
 'nr',
 'psx',
 'r_values',
 'rho_buf',
 'shifts',
 'sn',
 'sv',
 'vr']

We can adjust which we visualize using the following syntax.

In [5]:

qm.component_list = ["r_values", "SNR_comp", "cnn_preds", "neurons_sn"]
noise = qm.components.neurons_sn
qm.plots = {
    "neurons_sn": {
        "xaxis": "Noise Std",
        "data": qm.components.neurons_sn,
        "bins": np.linspace(min(noise), max(noise), 10),
        "vline": noise.mean(),
    }
}
fig = qm.get_grid(n_columns=4)
fig.show("png")

qm.component_list = ["r_values", "SNR_comp", "cnn_preds", "neurons_sn"] noise = qm.components.neurons_sn qm.plots = { "neurons_sn": { "xaxis": "Noise Std", "data": qm.components.neurons_sn, "bins": np.linspace(min(noise), max(noise), 10), "vline": noise.mean(), } } fig = qm.get_grid(n_columns=4) fig.show("png")

In [6]:

qm.components

qm.components

Out[6]:

	r_values	SNR_comp	cnn_preds	neurons_sn
0	0.500398	1.863260	0.045389	5.313188
1	0.568861	1.033131	0.031540	8.824791
2	0.226204	0.618304	0.012997	9.883841
3	0.599668	1.234195	0.025490	8.781350
4	0.000000	0.000000	0.000000	0.042582
...	...	...	...	...
108	0.652990	1.495659	0.000519	13.064044
109	0.765022	1.498043	0.001978	7.738464
110	0.803198	0.420331	0.002988	10.223336
111	0.588237	1.763111	0.033076	8.086799
112	0.175811	1.353125	0.008852	11.129535

113 rows × 4 columns

In [ ]: