Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
* Squashed commit of the following: commit 8564eb0 Merge: f6ef62c 07cc7d8 Author: Kendell Clement <k.clement.dev@gmail.com> Date: Tue Jan 11 16:20:15 2022 -0500 Merge branch 'indel-alignment-fix' of https://github.com/edilytics/CRISPResso2 into indel-alignment-fix commit 07cc7d8 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:29:59 2021 -0700 Fix bug in `find_indels_substitutions` This bug occurred when there was a deletion at the end of a sequence, and was thus not properly accounted for. commit f6ef62c Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:29:59 2021 -0700 Fix bug in `find_indels_substitutions` This bug occurred when there was a deletion at the end of a sequence, and was thus not properly accounted for. commit 7212f87 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:26:17 2021 -0700 Add a unit test for `find_indels_substitutions` This unit test checks for deletions at the end of a sequence, which are inherently outside of the include_indx_set window. commit d50b4e9 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:03:22 2021 -0700 Fix a bug in `find_indels_substitutions` The bug that this commit fixes is when an insertion occurs at the edge of the include indexes. The trouble with this earlier was that it was using the `idx` to calculate the size of the insertion, but the `idx` wasn't being incremented anymore because it was outside of the include window. commit 4db066f Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:01:39 2021 -0700 Add test case for `find_indels_substitutions` This test case is extracted from the CRISPRessoBatch integration test and provides an example where there is an insertion at the edge of the include index. commit 3b3a741 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 11:37:07 2021 -0700 Fix bug in CRISPRessoCompare where sample names were not properly set This was a place where it was (partially) missed during the crispresso2_info object refactoring. commit e9f5eff Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:26:17 2021 -0700 Add a unit test for `find_indels_substitutions` This unit test checks for deletions at the end of a sequence, which are inherently outside of the include_indx_set window. commit d4d45a9 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:03:22 2021 -0700 Fix a bug in `find_indels_substitutions` The bug that this commit fixes is when an insertion occurs at the edge of the include indexes. The trouble with this earlier was that it was using the `idx` to calculate the size of the insertion, but the `idx` wasn't being incremented anymore because it was outside of the include window. commit 13f00bb Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 15:01:39 2021 -0700 Add test case for `find_indels_substitutions` This test case is extracted from the CRISPRessoBatch integration test and provides an example where there is an insertion at the edge of the include index. commit 659ae34 Author: Cole Lyman <cole@colelyman.com> Date: Fri Dec 10 11:37:07 2021 -0700 Fix bug in CRISPRessoCompare where sample names were not properly set This was a place where it was (partially) missed during the crispresso2_info object refactoring. * Add parameter `--suppress_batch_summary_plots` If many runs are run at the same time, batch summary plots may fail because they are too large for matplotlib. This parameter `--suppress_batch_summary_plots` allows individual runs to be plotted, but suppresses batch summary plots that may otherwise be too big. * Pep formatting cleanup * Add summary nucleotide plots to aggregate * Aggregate plots are paginated * Update CRISPRessoAggregateCORE.py Remove max sample limit for plotting * Add --max_samples_per_summary_plot to CRISPRessoAggregate Parameterize the max number of samples to plot on each page of reports. Additional PDFs will be created with this number of samples on them. * Add plotly function to plot an interactive heatmap * Fix deprecated numpy type to suppress warning * Add plotting of heatmaps to CRISPRessoAggregateCORE to summarize modification types These heatmaps are interactive (zoomable and panable) and show for each sample the percentage of insertions, substitutions, and deletions. * Add the heatmap summaries to the CRISPRessoAggregate report * Update Bootstrap to 5.1.3 This is mainly so that we can use the fullscreen modal functionality in this version. * Move the plotly heatmaps to a Bootstrap modal * Fix bug where plots were not filling up entire modal. I have tried countless different ways for this to work, and this is the best that I can come up with. After the modal is opened it triggers the plot to resize, and then for some reason you need to trigger the resize event. I think this is because a `div` changing size won't actually trigger the resizing of the plot (and neither will just calling `Plotly.Plots.resize`...?!). * Update the axis labels and add autosize to plotly heatmaps I'm pretty sure the autosize doesn't do anything, but it is there for good measure. * Abandon attempts to make plots fullscreen This includes removing the Bootstrap modal (two out of the three plots would resize properly and I couldn't figure out a way to have the plot displayed outside of the modal). I have left in some javascript to make the plot fullscreen, but I couldn't get the formatting quite right and the plot wasn't much bigger in the fullscreen version because there was a ton of space between the plot and the heatmap. If some brave soul would like to tackle it, feel free! * Rename and refactor how plot data is passed around I have consolidated how the plot data is passed around, so that now you can pass in only one dict with all of the information instead of 4 or 5 separate parameters. I also renamed the `heatmap_plot_*` to `allele_modification_heatmap_*`. * Implement the line plot version of the modification percentages This also includes correctly resizing the plot when the line plot tab is selected! * Change default `max_samples_per_summary_plot` to be 150 instead of 250 * Remove extra assignments of `this_number_samples` and suppress plot The plot that is suppressed is the large nucleotide quilt when there is a large number of samples. Is it okay to suppress this plot @kclem? * Implement parallel plotting in CRISPRessoAggregate * Fix sample indexing error and heatmap scaling for large number of samples * Add parameter `--suppress_batch_summary_plots` If many runs are run at the same time, batch summary plots may fail because they are too large for matplotlib. This parameter `--suppress_batch_summary_plots` allows individual runs to be plotted, but suppresses batch summary plots that may otherwise be too big. * Pep formatting cleanup * Add summary nucleotide plots to aggregate * Aggregate plots are paginated * Update CRISPRessoAggregateCORE.py Remove max sample limit for plotting * Add --max_samples_per_summary_plot to CRISPRessoAggregate Parameterize the max number of samples to plot on each page of reports. Additional PDFs will be created with this number of samples on them. * Add plotly function to plot an interactive heatmap * Fix deprecated numpy type to suppress warning * Add plotting of heatmaps to CRISPRessoAggregateCORE to summarize modification types These heatmaps are interactive (zoomable and panable) and show for each sample the percentage of insertions, substitutions, and deletions. * Add the heatmap summaries to the CRISPRessoAggregate report * Update Bootstrap to 5.1.3 This is mainly so that we can use the fullscreen modal functionality in this version. * Move the plotly heatmaps to a Bootstrap modal * Fix bug where plots were not filling up entire modal. I have tried countless different ways for this to work, and this is the best that I can come up with. After the modal is opened it triggers the plot to resize, and then for some reason you need to trigger the resize event. I think this is because a `div` changing size won't actually trigger the resizing of the plot (and neither will just calling `Plotly.Plots.resize`...?!). * Update the axis labels and add autosize to plotly heatmaps I'm pretty sure the autosize doesn't do anything, but it is there for good measure. * Abandon attempts to make plots fullscreen This includes removing the Bootstrap modal (two out of the three plots would resize properly and I couldn't figure out a way to have the plot displayed outside of the modal). I have left in some javascript to make the plot fullscreen, but I couldn't get the formatting quite right and the plot wasn't much bigger in the fullscreen version because there was a ton of space between the plot and the heatmap. If some brave soul would like to tackle it, feel free! * Rename and refactor how plot data is passed around I have consolidated how the plot data is passed around, so that now you can pass in only one dict with all of the information instead of 4 or 5 separate parameters. I also renamed the `heatmap_plot_*` to `allele_modification_heatmap_*`. * Implement the line plot version of the modification percentages This also includes correctly resizing the plot when the line plot tab is selected! * Change default `max_samples_per_summary_plot` to be 150 instead of 250 * Remove extra assignments of `this_number_samples` and suppress plot The plot that is suppressed is the large nucleotide quilt when there is a large number of samples. Is it okay to suppress this plot @kclem? * Implement parallel plotting in CRISPRessoAggregate * Fix sample indexing error and heatmap scaling for large number of samples * Add plotly requrement to setup.py * Remove space around vertical barcharts * Add scrollbar to long images in multiReport * Fill in default (empty) values to allele modification plots When not running CRISPRessoAggregate, default values for the `allele_modification_heatmap_plot` and `allele_modification_lin_plot` dictionaries will be set so that the template can be properly rendered. * Include CRISPRessoBatch in the refactor of how summary_plot dicts are handled * Update dockerfile for new docker * minor bug fixes for plotCustomAllelePlot.py to work with Python3 (#212) * Allow for flexible parsing of quant window coordinates * CRISPRessoPooled debug flash command, fix pep formatting * Set flexiguide homology parameter type to int * Coerce ints in batch file checking (#200) * Batch type coerce and r2 file check * Revert "Batch type coerce and r2 file check" This reverts commit f917366. * Coerce int values * Handle multiple qwcs in batch mode If multiple qwcs were provided in batch mode, a parsing error would occur. This fixes this bug. * Fix bug from old pandas for int cols Evidently old pandas versions throw an error if a column doesn't exist. This checks to see if the column exists before the values are set. * Create allele modification heatmaps and line plots in CRISPRessoBatch * Add allele modification heatmaps and line plots to CRISPRessoBatch * Make all plots in CRISPRessoBatch run in parallel * Make `--suppress_batch_summary_plots` store true Also, only open and shutdown the process pool when necessary. * Add blank values for allele_modification entries when not present Co-authored-by: Kendell Clement <k.clement.dev@gmail.com> Co-authored-by: dharjanto <dewi.harjanto@gmail.com> Co-authored-by: Samuel Nichols <Snic9004@gmail.com>
- Loading branch information
4 people
authored
May 4, 2022
1 parent
f67376f
commit 62900e9