| _version_ | 1866902042050560000 |
|---|---|
| author | Playout, Clément Mehrjoo, Yosra Duval, Renaud Boucher, Marie Carole Costantino, Santiago Wurtele, Hugo |
| author_facet | Playout, Clément Mehrjoo, Yosra Duval, Renaud Boucher, Marie Carole Costantino, Santiago Wurtele, Hugo |
| contents | <p>Data corresponding to the project:</p> <p>DNAi: an open-source AI tool for unbiased DNA fiber analysis</p> <p>The dataset used for training and evaluation of <strong>DNAi</strong> consists of <strong>1,032 non-overlapping tiles</strong> of immunofluorescence microscopy images of DNA fibers, obtained after incubation with CldU (red channel), IdU (green channel), and spreading.</p> <ul> <li> <p><strong>Training set:</strong> 831 images with manual annotations</p> </li> <li> <p><strong>Test set:</strong> 201 images with manual annotations</p> </li> <li> <p><strong>Inter-grader study:</strong> 20 images (subset of the test set) manually annotated independently by four expert annotators, to assess inter-annotator variability.</p> </li> </ul> <p>All images were acquired using a <strong>Zeiss Axio Observer Z2 widefield fluorescence microscope</strong> equipped with a <strong>63×/1.4 oil immersion objective lens</strong> and a <strong>Zeiss Axiocam 820 monochrome sCMOS camera</strong>. The pixel size is <strong>0.26 μm × 0.26 μm</strong>.</p> <p>Abstract:<br>DNA fiber assays are powerful tools for investigating replication dynamics at the single-molecule level. However, their application and widespread adoption has been hampered by the labor-intensive and tedious nature of manual analysis of large numbers of images. Quantification of labeled DNA fibers typically depends on subjective examination, selection, and annotation of individual fibers from fluorescence microscopy images reducing inter-user consistency, reproducibility, and experimental throughput. To address these issues, we developed DNAi, a computer vision tool based on deep learning allowing automated detection and quantification of labeled DNA fiber length. DNAi was trained on a large and diverse dataset of manually annotated images of DNA fibers and matches human performance and accuracy in segmentation and length measurement across a wide range of experimental conditions. The open-source tool includes a user-friendly interface, which permits visual validation and manual selection of segmented fibers. Overall, DNAi enables robust, rapid, and reproducible DNA fiber analysis, and is freely available.</p> <div> </div> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17237308 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | DNAi: an open-source AI tool for unbiased DNA fiber analysis Playout, Clément Mehrjoo, Yosra Duval, Renaud Boucher, Marie Carole Costantino, Santiago Wurtele, Hugo DNA fibers analyses Replicative stress DNA replication DNA spreading Deep learning <p>Data corresponding to the project:</p> <p>DNAi: an open-source AI tool for unbiased DNA fiber analysis</p> <p>The dataset used for training and evaluation of <strong>DNAi</strong> consists of <strong>1,032 non-overlapping tiles</strong> of immunofluorescence microscopy images of DNA fibers, obtained after incubation with CldU (red channel), IdU (green channel), and spreading.</p> <ul> <li> <p><strong>Training set:</strong> 831 images with manual annotations</p> </li> <li> <p><strong>Test set:</strong> 201 images with manual annotations</p> </li> <li> <p><strong>Inter-grader study:</strong> 20 images (subset of the test set) manually annotated independently by four expert annotators, to assess inter-annotator variability.</p> </li> </ul> <p>All images were acquired using a <strong>Zeiss Axio Observer Z2 widefield fluorescence microscope</strong> equipped with a <strong>63×/1.4 oil immersion objective lens</strong> and a <strong>Zeiss Axiocam 820 monochrome sCMOS camera</strong>. The pixel size is <strong>0.26 μm × 0.26 μm</strong>.</p> <p>Abstract:<br>DNA fiber assays are powerful tools for investigating replication dynamics at the single-molecule level. However, their application and widespread adoption has been hampered by the labor-intensive and tedious nature of manual analysis of large numbers of images. Quantification of labeled DNA fibers typically depends on subjective examination, selection, and annotation of individual fibers from fluorescence microscopy images reducing inter-user consistency, reproducibility, and experimental throughput. To address these issues, we developed DNAi, a computer vision tool based on deep learning allowing automated detection and quantification of labeled DNA fiber length. DNAi was trained on a large and diverse dataset of manually annotated images of DNA fibers and matches human performance and accuracy in segmentation and length measurement across a wide range of experimental conditions. The open-source tool includes a user-friendly interface, which permits visual validation and manual selection of segmented fibers. Overall, DNAi enables robust, rapid, and reproducible DNA fiber analysis, and is freely available.</p> <div> </div> |
| title | DNAi: an open-source AI tool for unbiased DNA fiber analysis |
| topic | DNA fibers analyses Replicative stress DNA replication DNA spreading Deep learning |
| url | https://doi.org/10.5281/zenodo.17237308 |