yo-aka-gene / algebra_ver822

Analysis codes for "A set-theoretic definition of cell types with an algebraic structure on gene regulatory networks and application in annotation of RNA-seq data"

algebra_ver822's Issues

Split downloaded files from GSE and rename files

• Unless rename files to barcodes.tsv.gz/features.tsv.gz/matrix.mtx.gz, Seurat doesn't recognize them
• if the group of files have multiple prefix such as xxx.barcodes.tsv.gz and yyy.barcodes.tsv.gz, we have to move them to /xxx/barcodes.tsv.gz and /yyy/barcodes.tsv.gz respectively

Invalid Filename and Invalid Headline in ./data/m1_10x_raw

What

• run fmt_mtx again
• run fmt_tsv with alias (features -> genes)

Makefile or Dockerfile to share versions of pkgs between Jupyter and Rstudio

What

• To share codes between Jupyter and Rstudio, we need to share information of requirements_py.txt and requirements_r.csv

Colnames and rownames of dgCMatrix should be exported as single .json file

What

• In data preprocessing via Seurat, colnames and rownames are exported separately as .txt files
• For the sake of consistency in data loading by spongy_panda, they should be expoerted as single .json file
• Maybe it's better to implement a func which directly forms SparseDF from .mtx and .json file

Try implement to concatenate `.mtx` files directly to conjugate separated datasets

Backgrounds

• Datasets are separated in multiple subsets (e.g., m1_10x and GSE165288)
• For the huge data sizes, it is nearly impossible to concatenate them after they are loaded even as sparse matrices

What

• Therefore, we need to try another approach
• We can directly edit .mtx files and barcodes.tsv files

raw data preprocess of m1_10x in .mtx format

What

• raw data of m1_10x can be formatted as .mtx so that we can use preprocessing pipeline of seurat

Add slices with list in spongy_pandas.SparseDF

What

• to treat like pd.DataFrame, at least loc and iloc is necessary
• slices with list (of integers or strings) is also needed
• maybe we need to split definition of base class and surface class (like pd.DataFrame and pd.core.frame.DataFrame)

Open .npz file in R studio

Why

• some automatic annotation packages are only available in R
• therefore when I try to compare results of annotation (based on GRN structure) with the results of automated annotation, I need to open RNA-seq data (at least m1_10x) in R studio
• As m1_10x is super massive data and the resource of memory is limited for R studio, I need to convert file format to .npz to compress
• accordingly, I need to install numpy in R studio

Procedure

• install RStudio Package Manager. See also the Admin Guide
• mirror PypI
• install numpy
• add commands in Dockerfile or Makefile to assure reproductivity

Conjugate GW9 to GW12 in data preprocessing section for factor analysis in python

What

• submatrices in GSE165388 were processed by Seurat
• Although, they should be concatenated for further analysis in python

Ignore .txt or .json files in data/gse165388_*/

What

• Data preprocessing via Seurat exports .mtx file and .txt file (in the future .json file) to enable python codes to read processed data
• As the .json file are generated as many as the number of submatrices of the data (.txt data are generated twice as many as the submatrices), they should be removed

Try larger dataset to embed in tSNE in visualization of m1_10x

• The number of explanatory genes seemed quite small to visualize cellular diversities in current impelmetnation

🚨 failed to run `docker compose up -d`

Error response from daemon: failed to create shim: OCI runtime create failed: container_linux.go:380: starting container process caused: exec: "start.sh": executable file not found in $PATH: unknown