Wiki source code of HBP Summit 2023 - Atlas pipeline workflow

Version 19.1 by drodarie on 2023/03/26 17:05

version	line-number	content
1.1	1	(% class="jumbotron" %)
	2	(((
	3	(% class="container" %)
	4	(((
6.2	5	= Creating a mouse cell atlas model =
1.1	6
6.2	7	From annotation displaying to neuron density estimations
12.1	8
	9	(% style="text-align:center" %)
	10	[[image:workflow_complete_plot.jpg]]
1.1	11	)))
	12	)))
	13
	14	(% class="row" %)
	15	(((
	16	(% class="col-xs-12 col-sm-8" %)
	17	(((
14.1	18	= Content =
1.1	19
17.2	20	This collab is meant for modelists interested in using [[Blue Brain atlas suite>>https://github.com/BlueBrain/atlas-suite]] [1] and the [[Brain Scaffold Builder>>https://github.com/dbbs-lab/bsb]] (BSB) [2] to build circuits of the mouse brain.
1.1	21
14.1	22	This collab contains notebooks to present the mouse brain cell atlas pipeline: its steps and results.
	23
	24	Through these notebooks, you should learn:
	25
17.2	26	* How to manipulate the Allen Brain reference atlases and brain region hierarchy, using the [[voxcell>>https://github.com/BlueBrain/voxcell]] library.
	27	* (% style="color:#999999" %)How to realign ISH datasets to the Nissl reference atlas (introduction to [[Deep-Atlas>>https://github.com/BlueBrain/Deep-Atlas]] [3]).
	28	* How to compute the orientation field followed by the neurons inside a region (introduction to [[atlas-direction-vectors>>https://github.com/BlueBrain/atlas-direction-vectors]]).
14.1	29	* (% style="color:#999999" %)How to compute the depth field following the orientation field.
17.2	30	* How to compute cell, neuron and glia densities in each region of the mouse brain (first introduction to [[atlas-densities>>https://github.com/BlueBrain/atlas-densities]]).
	31	* (% style="color:#999999" %)How to estimate neuron sub-type densities based on literature findings (second introduction to (%%)[[atlas-densities>>https://github.com/BlueBrain/atlas-densities]](% style="color:#999999" %)).
14.1	32	* (% style="color:#999999" %)How to refine neuron densities based on linear densities.
	33	* (% style="color:#999999" %)How to scale, rotate and bend neuron morphologies, based on the orientation and depth fields.
17.3	34	* (% style="color:#999999" %)How to construct a circuit based on the resulting cell atlas model in [[BSB>>https://github.com/dbbs-lab/bsb]].
13.1	35
14.1	36	Each item in the list correspond to a notebook in the lab. Items in grey are notebooks still in preparation and will be integrated in the HBP collab, as soon as possible.
1.1	37
13.1	38
14.1	39	= Jupyter notebooks =
13.1	40
14.1	41	Inside the lab, you should find notebooks numbered so that they follow the natural steps of the pipeline.
13.1	42
14.1	43	Some parts of the notebooks might require more than the 2GB RAM memory available on the HBP lab. These steps will be highlighted in the notebook.
	44
19.1	45
14.1	46	= About the figure. =
	47
13.1	48	Figure 1. Mouse declive cell placement workflow:
	49
	50	Each image represents a sagittal slice of the result of the step of our process to reconstruct the mouse declive. Black arrows show the order of the workflow steps.
	51
	52	A. Our corrected annotation atlas of declive is shown in colors over the Nissl volume in levels of grey. The granular layer appears in orange, the molecular layer in blue and Purkinje layer in green.
	53
	54	B. To each voxel of the declive, a 3D direction normalized vector is computed corresponding to the main axis of the axons in the region. Colors represent the orientation vectors norm on their respective plane, black arrows their projected vector.
	55
	56	C. Distance of each voxel of declive to the outside border of the molecular layer, following the orientation field, expressed in micrometers.
	57
	58	D. E. Respectively Neuron and inhibitory neuron density in logarithmic scale.
	59
	60	F. Soma position of the different neuron types of the declive, displayed over the annotation atlas. Each cell type appears in a different color and size corresponding to its radius. The annotation atlas’ colors correspond to A.
	61
	62	G. Projection of the Purkinje cell morphology displayed in colors over the annotation atlas. Each morphology has been rotated, scaled, and bended following the orientation and depth fields. The annotation atlas’ colors correspond to A.
14.1	63
	64
	65	= References. =
	66
	67
	68	1. D. Rodarie //et al.//, “A method to estimate the cellular composition of the mouse brain from heterogeneous datasets,” //PLoS Comput Biol//, vol. 18, no. 12, p. e1010739, Dec. 2022, doi: [[10.1371/journal.pcbi.1010739>>url:https://doi.org/10.1371/journal.pcbi.1010739]].
	69	1. (((
	70	R. De Schepper //et al.//, “Model simulations unveil the structure-function-dynamics relationship of the cerebellar cortical microcircuit,” //Commun Biol//, vol. 5, no. 1, Art. no. 1, Nov. 2022, doi: [[10.1038/s42003-022-04213-y>>url:https://doi.org/10.1038/s42003-022-04213-y]].
1.1	71	)))
14.1	72	1. (((
	73	J. Krepl //et al.//, “Supervised Learning With Perceptual Similarity for Multimodal Gene Expression Registration of a Mouse Brain Atlas,” //Front. Neuroinform.//, vol. 15, p. 691918, Jul. 2021, doi: [[10.3389/fninf.2021.691918>>url:https://doi.org/10.3389/fninf.2021.691918]].
	74	)))
	75	)))
1.1	76
	77
	78	(% class="col-xs-12 col-sm-4" %)
	79	(((
	80	{{box title="Contents"}}
	81	{{toc/}}
	82	{{/box}}
	83
	84
	85	)))
	86	)))