Cultures of healthy and ALS iPSCs-derived motor neurons immunolabeled for Beta Tubulin III and TDP-43 and counterstained with DAPI

The archive (fluoMNs_subset.zip) contains five folders corresponding to five screens: E, F, G, H and I. Each screen has one or several folders corresponding to one or several plates. The content of the archive is annotated in annotations_fluoMNs_subset.csv. Each row corresponds to one field of fiew in the 96-well plates. Each well in the plate has 10 to 12 fields of view, each field of view has 4 to 5 planes and each plane was stained with 3 to 4 channels. Each image is encoded as follows: 00R00C-F-00100P00H, where R is the row of the well in the plate, C is the column of the well in the plate, F is the field of view, P is the plane and H is the channel.

Here are two Python functions to get the complete path to one image:

def get_filename(source_directory, experiment: str, plate: str, image_id: str) -> str:
   """Return pathname of an image"""
   return f'{source_directory}/{experiment}/{plate}/{image_id}.tif'

def get_image_id(row: int, col: int, fov: int, plane: int, channel: int) -> str:
   """Return id of an image"""
   return f'{row:03}{col:03}-{fov}-001{plane:03}{channel:03}'

This is a subset of a larger dataset of cultures of motor neurons under different celullar stress and immunolabeled for ALS-linked RNA-binding proteins. The whole dataset (~1TB) is available on IDR.

Some code using this data subset for classification of ALS vs healthy motor neurons can be accessed on Github.

This data has been used in a manuscript posted on BioRxiv.