Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from https://github.com/Lujun995/DiSC_Replication_Code below:

GitHub - Lujun995/DiSC_Replication_Code

This folder contains scripts, data, and R markdown files that can be used to reproduce the results presented in our paper. In the main directory, you will find four markdown files along with their corresponding HTML outputs: Real_data_simulation.html, Parametric_simulation.html, Real_data_analysis.html, Supplementary_information.html, and Response_to_Reviewers.html. We recommend starting a code review with these R markdown outputs in HTML format. Additionally, there are several subfolders for organizing data, source code, scripts, parameters, intermediate results, and more.

Cautionary Note: In this replication code, the feature "nzm" used in the DiSC function corresponds to the sample mean. In our package SingleCellStat (version 0.3.1), this has been corrected to "m" for clarity. However, to maintain backward compatibility, the replication code continues to use "nzm" to represent the sample mean while the DiSC function in this context does not accept "m" as a feature.

• R Markdown: Real_data_simulation.Rmd
• R Markdown Output: Real_data_simulation.html
• Corresponding Data Folder: real_data_sim
• Zipped files: type1_0715.zip.001 and type1_0715.zip.002 need to be unzipped before use.
• Description: These simulations are based on cell type "L2/3" in the prefrontal cortex (PFC) region, sourced from an autism study conducted by Velmeshev, et al. (2019). We permuted labels for the variable diagnosis (ASD vs. control) and applied various statistical methods to investigate the association between diagnosis and single-cell RNA sequencing data, with a focus on examining type I error rates and false discovery rates.

• R Markdown: Parametric_simulation.Rmd
• R Markdown Output: Parametric_simulation.html
• Corresponding Data Folder: para_data_sim
• Zipped files: results_n12_12_d1_c375_vFALSE_0811.zip.001 and results_n12_12_d1_c375_vFALSE_0811.zip.002 need to be unzipped before use.
• Description:
  • The parametric simulation begins with parameters estimated from real-world data, derived from the autism study by Velmeshev et al. (2019) using a deep count autoencoder (DCA) as described by Eraslan et al. (2019) at the cell level. These estimations are then aggregated at the individual level for the simulation.
  • The parametric simulation encompasses four scenarios and three types of changes. The four scenarios include increased effect sizes, increased cell numbers, increased sample sizes, and a no-signal scenario designed to examine type I error and false discovery rate (FDR). The three types of changes are differences in mean, variance, and both mean and variance.
  • The primary objective is to assess the statistical power, type I error rate, and false discovery rate of different methods.

• R Markdown: Real_data_analysis.Rmd
• R Markdown Output: Real_data_analysis.html
• Corresponding Data Folder: real_data_anal
• Description: The data used in these analyses come from 17 distinct cell types within the prefrontal cortex region, as part of the autism study by Velmeshev et al. (2019). These analyses aim to test the association between autism spectrum disorder (ASD) and single-cell RNA sequencing data, while adjusting for covariates such as age, sex, sequence batch (Seqbatch), and RNA Integrity Number (RIN). The analyses employ various methods, and the results are compared.

• R Markdown: Supplementary_information.rmd
• R Markdown Output: Supplementary_information.html
• Corresponding Data Folder: supp_info_anal
• Description: The example data set used for these analyses was generated in the "Simulations Based on Parametric Models" step. It includes 8,000 genes and 12 cases and 12 controls, each with 375 cell replicates. The objectives are to (i) determine the number of permutations required for DiSC; (ii) compare the computational time of different methods; (iii) evaluate memory usage by DiSC; and (iv) examine the impact of dropping a proportion of insufficiently sequenced cells during the rarefaction step on the results.

• R Markdown: Response_to_Reviewers.rmd
• R Markdown Output: Response_to_Reviewers.html
• Corresponding Data Folder: res_to_rev
• Description: The code and scripts used to address the reviewers' concerns and incorporate their suggestions. Results generated in this part are annotated with "_rev" in the results folder.

• ideas_pipeline-main: This folder contains scRNA-seq data from the prefrontal cortex (PFC) region of both autism patients and healthy controls, sourced from Velmeshev et al. (2019). The folder also hosts source code necessary for the DCA-IDEAS method, as discussed in Zhang et al. (2022). These contents were adapted from the ideas_pipeline by Zhang et al. (2022).
• results: This directory houses figures and tables generated from the R markdown files above, which are included in our manuscript.
• source_code: This folder contains the source code needed to reproduce the tables and figures from the paper. Please note that these files cannot be directly executed, as they require additional parameter specifications before running. The complete scripts can be found in the folders mentioned above: real_data_sim/, para_data_sim/, real_data_anal/, and supp_info_anal.

• Zhang, L., Yang, L., Ren, Y., Zhang, S., Guan, W., & Chen, J. (Submitted to Bioinformatics): DiSC: a Statistical Tool for Fast Differential Expression Analysis of Individual-level Single-cell RNA-seq Data.
• Eraslan, G., et al. Single-cell RNA-seq denoising using a deep count autoencoder. Nature communications, 10.1 (2019): 1-14.
• Velmeshev, D., et al. Single-cell genomics identifies cell type–specific molecular changes in autism. Science, 364.6441 (2019): 685-689.
• Zhang, M., et al. IDEAS: individual level differential expression analysis for single-cell RNA-seq data. Genome Biol, 23, 33 (2022). DOI.

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4