2024-03-28T17:29:09Zhttp://repisalud.isciii.es/oai/requestoai:repisalud.isciii.es:20.500.12105/71422022-11-03T19:21:15Zcom_20.500.12105_2152com_20.500.12105_2051com_20.500.12105_2144com_20.500.12105_2145col_20.500.12105_2153col_20.500.12105_2146
Repisalud
author
Jimenez-Carretero, Daniel
author
Abrishami, Vahid
author
Fernandez-de-Manuel, Laura
author
Palacios, Irene
author
Quilez-Alvarez, Antonio
author
Diez-Sanchez, Alberto
author
del Pozo, Miguel Angel
author
Montoya, Maria
funder
Ministerio de Ciencia, Innovación y Universidades (España)
funder
Unión Europea. Comisión Europea
funder
Fundación ProCNIC
funder
Worldwide Cancer Research
funder
Fundación La Caixa
2019-02-07T13:51:09Z
2019-02-07T13:51:09Z
2018
PLoS Comput Biol. 2018; 14(11):e1006238
1553-7358
http://hdl.handle.net/20.500.12105/7142
30500821
10.1371/journal.pcbi.1006238
1553-7358
PLoS computational biology
Toxicity is an important factor in failed drug development, and its efficient identification and prediction is a major challenge in drug discovery. We have explored the potential of microscopy images of fluorescently labeled nuclei for the prediction of toxicity based on nucleus pattern recognition. Deep learning algorithms obtain abstract representations of images through an automated process, allowing them to efficiently classify complex patterns, and have become the state-of-the art in machine learning for computer vision. Here, deep convolutional neural networks (CNN) were trained to predict toxicity from images of DAPI-stained cells pre-treated with a set of drugs with differing toxicity mechanisms. Different cropping strategies were used for training CNN models, the nuclei-cropping-based Tox_CNN model outperformed other models classifying cells according to health status. Tox_CNN allowed automated extraction of feature maps that clustered compounds according to mechanism of action. Moreover, fully automated region-based CNNs (RCNN) were implemented to detect and classify nuclei, providing per-cell toxicity prediction from raw screening images. We validated both Tox_(R)CNN models for detection of pre-lethal toxicity from nuclei images, which proved to be more sensitive and have broader specificity than established toxicity readouts. These models predicted toxicity of drugs with mechanisms of action other than those they had been trained for and were successfully transferred to other cell assays. The Tox_(R)CNN models thus provide robust, sensitive, and cost-effective tools for in vitro screening of drug-induced toxicity. These models can be adopted for compound prioritization in drug screening campaigns, and could thereby increase the efficiency of drug discovery.
eng
Tox_(R)CNN: Deep learning-based nuclei profiling tool for drug toxicity screening
journal article
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
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/1/Tox_%28R%29CNNDeepLearning-Based_2018.pdf
File
MD5
990262b76c84ffe09155e7aaed3fa90e
3136878
application/pdf
Tox_(R)CNNDeepLearning-Based_2018.pdf
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/2/pcbi.1006238.s002.tif
File
MD5
926064274d358c486775f54ad459a54c
1452804
image/tiff
pcbi.1006238.s002.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/3/pcbi.1006238.s001.tif
File
MD5
7bee27c682833756c20416a224d414a1
5658714
image/tiff
pcbi.1006238.s001.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/4/pcbi.1006238.s004.tif
File
MD5
1bf644dd79b127487b6f6fed07cd5e5e
976886
image/tiff
pcbi.1006238.s004.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/5/pcbi.1006238.s003.tif
File
MD5
0935c9fcddf18410f18eced56b33ad0f
1823022
image/tiff
pcbi.1006238.s003.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/6/pcbi.1006238.s007.xlsx
File
MD5
55653ab7ed384dd9b8604e99169741c5
11881
application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
pcbi.1006238.s007.xlsx
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/7/pcbi.1006238.s006.tif
File
MD5
4e46f603e475ac4b64a2f7efb9baef92
4974126
image/tiff
pcbi.1006238.s006.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/8/pcbi.1006238.s005.tif
File
MD5
ad23ded3693f28d2d8802f76e4df02d2
2078302
image/tiff
pcbi.1006238.s005.tif
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/9/pcbi.1006238.s010.zip
File
MD5
592f80e4b77f35c7a169618ce30122a1
888895
application/octet-stream
pcbi.1006238.s010.zip
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/10/pcbi.1006238.s009.xlsx
File
MD5
1cc0626bca4a1d6fd178d6d2675f8a8a
14417
application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
pcbi.1006238.s009.xlsx
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/10/pcbi.1006238.s008.xlsx
File
MD5
db86311e786382a68d2c309fabff6dc6
11314
application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
pcbi.1006238.s008.xlsx
URL
https://repisalud.isciii.es/bitstream/20.500.12105/7142/20/Tox_%28R%29CNNDeepLearning-Based_2018.pdf.txt
File
MD5
0579a6b93350f78e1e5ea67e4dd209ba
85773
text/plain
Tox_(R)CNNDeepLearning-Based_2018.pdf.txt