true barcode label

[DC-Jade]

A very nice tool
but How to determine the false reads and their matched true refs from simulate_main_umi() outputs, I do not find it, thank u

simulate_main_umi( )

[wenjie1991]

Hi @DC-Jade, thank you for your interest in the simulation tool.

Here is a brief example.

## install dependencies
install.packages("remotes")
BiocManager::install("DNABarcodes")

## Install CellBarcodeSim from GitHub
remotes::install_github("TeamPerie/CellBarcodeSim")

library(CellBarcodeSim)

barcode_library_file = system.file("data", "random_barcodes.tsv", package = "CellBarcodeSim")
head(read.table(barcode_library_file, header = T))

## Simulate a Barcode sequencing experiment
simulate_main(
    barcode_library_file = barcode_library_file,  ## Define the barcode library
    clone_n              = 20,  ## Define the number of clones
    clone_size_dist      = "uniform", ## Define the clone size distribution
    clone_size_dist_par  = list(size_max = 1000, size_min = 1),  ## Define the parameters of the clone size distribution
    cycle                = 30,  ## Define the number of cell cycles
    efficiency           = 0.705, ## Define the efficiency of the cell cycle
    error                = 1e-6,  ## Define the error rate of the PCR, mutation per base per cycle
    pcr_read_per_cell    = 50,  ## Define the number of PCR reads per cell (clone_n)
    output_prefix        = "./tmp/barcode_sim" ,  ## Define the output prefix
    ngs_profile          = "MSv1",  ## Define the NGS profile (refer to ART sequencing simulator manual)
    reads_length         = 35,  ## Define the length of the reads
    top_seq              = "AAAAAAAAAAGGGGG",  ## Define the fixed sequence at the 5' end of the reads to be added
    bottom_seq           = "TTTTTTTTTT",  ## Define the fixed sequence at the 3' end of the reads to be added
    sequence_trunk       = 10, ## Define the length of the fixed sequence to be added
    art_bin              = NULL ## Use the default ART binary file
)

There are four files:

1. User input barcode_library_file, defines the barcode library.
2. Output ./tmp/barcode_sim_ref.tsv file containing the induced barcodes and their corresponding frequencies in the sample.
3. Output ./tmp/barcode_sim_library_fasta file containing the sequencing library with PCR error.
4. Output ./tmp/barcode_sim.fq file containing the simulated sequencing FASTQ file.

Normally, you can get the "true induced barcode list" in ./tmp/barocde_sim_ref.tsv, this is the induced barcodes list. Which is different from the barcode library file `barcode_library_file.

However, the simulator does not track the "PCR mutations process" of the sequence. However, the sequences in ./tmp/barcode_sim_library_fasta and ./tmp/barcode_sim.fq match each other, allowing you to observe the sequencing mutation process.

We can modify the cpp code https://github.com/TeamPerie/CellBarcodeSim/blob/main/src/lib_pcr_simulation.cpp to enable to trace the "PCR mutation process". The only potential concern is the potential overhead.

Could you provide more information about your use case so we can assist you better?

[DC-Jade]

Thank you for your kind reply.
The reason I want to know the detailed labels is that there is a significant difference between the corrected read counts and the matched reference when I rerun the script at https://github.com/TeamPerie/CellBarcode_paper_Sun_et_al/blob/main/run_genBaRCode/run_test_genBaRcode.R.
I would appreciate it if you could add more information to the header of the FASTQ file—such as the matched original read, UMI, and barcode.
Regarding the "potential overhead" you mentioned, this could be set as an optional parameter to control the process

src: 20251023_run_test_genBaRcode.html

[wenjie1991]

Hi @DC-Jade,

I'm glad to hear you were able to re-run the results.

Firstly, I'm trying to understand the issue here.

There is a significant difference between the corrected read counts and the matched reference.

By "corrected read counts," are you referring to the corrected barcode frequency in the CellBarcode output?

barcode_seq count
1 TTCCTGAAGCGTCC 21932
307 GTTTCCCTTATGGA 18126
306 GCCGGTGACAAGTA 17517
305 CGTCGCAAATACTT 16798
304 TCATACGTGTCAAC 16579
303 GAAATATGCATGCT 12087
302 TGGGTGCTCGCGGT 11447
301 GTAGCTAATGGAAC 9298
300 AAGGCGAAAGATAA 8243
299 CGGCTTGACTGAGA 7467

And by "matched reference," do you mean the simulated barcode reference list?

barcode <- CellBarcodeSim::simu_barcode_random_uniform(length = barcode_len, n = 200000)

In the simulation, we perform the following steps separately:

1. Simulate the barcode library using simu_barcode_random_uniform(). You can see the barcode frequency in the barcode variable in the code above.

2. Simulate the biological clone size distribution, controlled by the parameters clone_size_dist and clone_size_dist_par in the CellBarcodeSim::simulate_main() function.

I suggest checking the simu_seq_ref.tsv file for the top frequent barcode "TTCCTGAAGCGTCC" in the "corrected read counts" table, as it should have a large clone size.

Please refer to the Result section in the attached Rmarkdown output.
20251023_run_test_genBaRcode_re.html

Please let me know if I understand your concern correctly, and I'll follow up here. Thanks.

[DC-Jade]

Hi Wenjie,

Your understanding is correct.
I mean while the corrected counts derived from CellBarcode/genBaRcode exhibit a strong correlation with the reference counts, they are not identical—as illustrated in the last two figures, where the corrected counts from CellBarcode/genBaRcode are higher than the reference counts.
Therefore, I would like to understand why certain barcodes cannot be corrected.

[wenjie1991]

I understand your request. It will need considerable effort to adding more information to the header of the FASTQ file, to quickly providing output to satisfy your specific interest, I recommend to add few lines of code to print all the PCR mutations. This will allow you to observe the mutation process during the PCR simulation. As for the sequencing simulation, you can identify sequencing errors by comparing the ./tmp/barcode_sim_library_fasta and ./tmp/barcode_sim.fq files.

By adding the following one line cpp code behind this line, you can track all the PCR mutation.

// string s_seq_new = mutate_seq(s_seq);
Rcout<<"----------------\nMutation event\noriginal_seq:"<<s_seq<<"\nmutant_seq:"<<s_seq_new<<"\n";

Here is an example of the output:
demo_print_pcr_mutation_process.html

I assume you're familiar with modifying and loading local R packages. If not, I can provide step-by-step instructions to update the package. Please let me know.

[DC-Jade]

I understand your request. It will need considerable effort to adding more information to the header of the FASTQ file, to quickly providing output to satisfy your specific interest, I recommend to add few lines of code to print all the PCR mutations. This will allow you to observe the mutation process during the PCR simulation. As for the sequencing simulation, you can identify sequencing errors by comparing the ./tmp/barcode_sim_library_fasta and ./tmp/barcode_sim.fq files.

By adding the following one line cpp code behind this line, you can track all the PCR mutation.

// string s_seq_new = mutate_seq(s_seq);
Rcout<<"----------------\nMutation event\noriginal_seq:"<<s_seq<<"\nmutant_seq:"<<s_seq_new<<"\n";

Here is an example of the output: demo_print_pcr_mutation_process.html

I assume you're familiar with modifying and loading local R packages. If not, I can provide step-by-step instructions to update the package. Please let me know.

I got it, thank you so much for your kind help!