Conversation
|
Click the link below to download the manuscript build as a ZIP file. There were 4 spelling errors. |
|
Since there's quite a bit of new text here, I'm going to request @jaclyn-taroni for primary review (with the understanding that it might take a minute to get to, which seems fine, we're on a good schedule here!). But, @allyhawkins if you also want to weigh in, in general or in particular on the questions about |
|
Click the link below to download the manuscript build as a ZIP file. There were 0 spelling errors. |
allyhawkins
left a comment
allyhawkins
left a comment
There was a problem hiding this comment.
I think the overall organization is good and I like talking about OpenScPCA first because then you can say that the inferCNV results match up with expectations based on OpenScPCA annotations. Because of that I would keep the figure order.
content/03.results.md
Outdated
| In parallel to developing the ScPCA Portal, we launched the OpenScPCA project [@url: https://openscpca.readthedocs.io], an open-science collaborative initiative to further characterize and analyze Portal data. | ||
| As part of this project, we have thus far performed manual cell type annotations on two projects, `SCPCP000004` (neuroblastoma) and `SCPCP000015` (Ewing sarcoma), and made the associated annotations available in the Portal for broad reuse. | ||
| Figure {@fig:fig5}A displays, for example, a UMAP of all libraries in `SCPCP000004` highlighting this project's OpenScPCA annotations. | ||
| These annotations were derived based on the `NBAtlas` reference [@doi:10.1016/j.celrep.2024.114804] using both `SingleR` and `scArches` [@doi:10.15252/msb.20209620; @doi:10.1038/s41587-021-01001-7], leveraging consensus cell type annotations to resolve differences between these methods; more information about the specific annotation procedure can be found in the `OpenScPCA-nf` GitHub repository at <https://github.com/AlexsLemonade/OpenScPCA-nf/tree/main/modules/cell-type-neuroblastoma-04>. <!-- TODO: Is this link overkill and adding too much text? We do link OpenScPCA-nf repo in Methods --> |
There was a problem hiding this comment.
If we are trying to keep things concise and shorter, then I agree that this is too much and we don't need to link it.
content/03.results.md
Outdated
| These annotations were derived based on the `NBAtlas` reference [@doi:10.1016/j.celrep.2024.114804] using both `SingleR` and `scArches` [@doi:10.15252/msb.20209620; @doi:10.1038/s41587-021-01001-7], leveraging consensus cell type annotations to resolve differences between these methods; more information about the specific annotation procedure can be found in the `OpenScPCA-nf` GitHub repository at <https://github.com/AlexsLemonade/OpenScPCA-nf/tree/main/modules/cell-type-neuroblastoma-04>. <!-- TODO: Is this link overkill and adding too much text? We do link OpenScPCA-nf repo in Methods --> | ||
| Unlike the consensus cell type annotations, the OpenScPCA project annotations distinguish between normal and malignant cells and contain far fewer uncharacterized cells. | ||
| Indeed, for `SCPCP000004`, the consensus cell type procedure labeled only ~43% of cells, but the OpenScPCA project labeled ~91% of cells, thereby adding substantial value to the data. | ||
| <!-- Calculation. TODO: Do we want this added to `scpca-paper-figures` manuscript-numbers folder? |
There was a problem hiding this comment.
I like having this number so I would say yes to this question.
Thanks for having a look, @allyhawkins! Noting that in my 1:1 with Jackie just now, we also agreed to keep the figure order so that's all set! I will be adding in a quick addition to |
Co-authored-by: Ally Hawkins <54039191+allyhawkins@users.noreply.github.com>
…pca annotations were done
|
Click the link below to download the manuscript build as a ZIP file. There were 0 spelling errors. |
jaclyn-taroni
left a comment
jaclyn-taroni
left a comment
There was a problem hiding this comment.
A couple comments, but I don't need to see this again
Co-authored-by: Jaclyn Taroni <19534205+jaclyn-taroni@users.noreply.github.com>
|
Click the link below to download the manuscript build as a ZIP file. There were 0 spelling errors. |
…results-section [ci skip] This build is based on 7e4d905. This commit was created by the following CI build and job: https://github.com/AlexsLemonade/ScPCA-manuscript/commit/7e4d9053d8d1d2619806ef331d98138dededdac8/checks https://github.com/AlexsLemonade/ScPCA-manuscript/actions/runs/20959556332
…results-section [ci skip] This build is based on 7e4d905. This commit was created by the following CI build and job: https://github.com/AlexsLemonade/ScPCA-manuscript/commit/7e4d9053d8d1d2619806ef331d98138dededdac8/checks https://github.com/AlexsLemonade/ScPCA-manuscript/actions/runs/20959556332
3 participants
Closes #214
This PR adds a (kind of hefty, but words can definitely be cut once we nail down the full content scope in review!) new section detailing the malignancy bits: limitations of consensus, openscpca annotations, and inferCNV. I structured text to follow the order of the figure 5 panels (for reference, here is figure 5) to introduce openscpca first, then infercnv, and finally how cnv + consensus can be used as a proxy to ID malignant cells. I think this flow is probably fine, but I also see a world where it's a bit funky and might slightly benefit from switching the panels 5A <-> 5B to be able to introduce inferCNV (a step embedded in
scpca-nf) before OpenScPCA (a step external toscpca-nf). This would result in a bit of text reorganization here, but I don't think that change is crucial. That repo is ready for release, but in case anyone wants to advocate hard for swapping that order I'm holding off on cutting the release just a smidge.A couple spots in particular to note:
manuscript-numbersfolder inscpca-paper-figures? Again, that repo is otherwise ready for release, but can add this in first if preferred.Just to summarize from the above, these are the questions for reviewer(s) that relate to
scpca-paper-figures: