journal_karbytes_01october2025
The text in the web page section below was copied from the plain-text file named journal_karbytes_01october2025_p0.txt in the “micro directory” website section named raw_github_files_directory_KARLINA_OBJECT_extension_pack_49 in the “macro directory” website section (named RAW_GITHUB_FILES_MACRO_DIRECTORY) of the website named Karbytes For Life Blog dot WordPress dot Com. Some of the aforementioned copied plain-text was then replaced with relevant hypertext on this web page.
To view hidden text inside of the preformatted text boxes below, scroll horizontally.
/**
* file: journal_karbytes_01october2025_p0.txt
* type: plain-text
* date: 01_OCTOBER_2025
* author: karbytes
* license: PUBLIC_DOMAIN
*/
Today I used the following Unix command to download an entire 983.7 megabyte compressed FASTA file of a reference human genome (so that I can practice archiving my own genome FASTA when it becomes tentatively available within the next six months):
[bash]
# UCSC Genome Browser (hg38 reference) wget http://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/hg38.fa.gz
[end bash]
Unforunately, the unzipped version of that file was too large to fit on my Desktop and caused my computer to freeze midway through expanding the zip file.
I inquired ChatGPT-5 about the difference between the HG38 Reference Genome FASTA file and the FASTA file I expect to be getting via Sequencing dot Com. Here is what ChatGPT-5 said:
[excerpt]
📊 Two very different “FASTA” situations 1. Reference genome FASTA (like UCSC hg38) - Contains every chromosome, contig, alternate locus in plain text. - That’s what you downloaded: hg38.fa.gz (~3 GB compressed, ~90–100 GB uncompressed). - This is the “master” reference and it’s huge. - Way too big to fit on a single 25 GB M-DISC unless you keep it compressed. 2. Personal genome FASTA (from sequencing companies) - After sequencing, many providers will give you a consensus FASTA of just your diploid genome sequence. - That’s ~3.2 billion bases (one byte per base) → about 3–6 GB uncompressed. - Compressed, often ~1 GB. - This easily fits on an M-DISC alongside your other data.
[end excerpt]
I thought it would be a good idea to archive the hg38.fa.gz compresssed FASTA file given that it can be used to “complete” my personal FASTA file.
I used the following Unix command to split the hg38.fa.gz file into chunks no larger than 20 megabytes each labeled with a unique integer starting with zero.
[bash]
split -b 20M -d --suffix-length=4 hg38.fa.gz hg38_chunk_
[end bash]
I uploaded one of those chunk files to a public GitHub repository I owned and tried saving its raw GitHub file URL to the WayBack Machine:
[test_url]
# Note that the uncensored version of this Uniform Resource Locator (URL) refers to one of my anonymous GitHub user accounts. https://raw.githubusercontent.com/[github_username]/[public_repository_name]/main/hg38_chunk_0000
[end test_url]
Apparently the 20mb file is too large to properly show up in the WayBack Machine. The saved web page ended up being a “404 page not found” result. The same error page result occurred when I tried saving MP4 videos to the WayBack Machine which contained too many non-identical frames.
I inquired ChatGPT-5 about it and ChatGPT-5 confirmed that “high entropy” files tend to not be saved to the WayBack Machine. I asked ChatGPT-5 if I should try chunking the FASTA into smaller sizes and it suggested 5 megabytes. Hence, I used the following Unix command to re-create the FASTA chunk files.
[bash]
split -b 5M -d --suffix-length=4 hg38.fa.gz hg38_chunk_
[end bash]
I uploaded one of those chunk files to a public GitHub repository I owned and tried saving its raw GitHub file URL to the WayBack Machine:
[test_url]
# Note that the uncensored version of this Uniform Resource Locator (URL) refers to one of my anonymous GitHub user accounts.
https://raw.githubusercontent.com/[github_username]/[public_repository_name]/main/hg38_chunk_0001
[end test_url]
It worked! I was able to save one of the ~5mb chunk files to the WayBack Machine such that the file can be downloaded from the WayBack Machine.
Next, I used the following command to combine all the chunked files into a copy of the original hg38.fa.gz file. The original was placed in a different file directory than the chunks to avoid that original file being overwritten.
[bash]
cat hg38_chunk_* > hg38.fa.gz
[end bash]
The result of running that command was getting a 983.7 megabyte file named hg38.fa.gz.
I generated a checksum of each FASTA file to ensure that both files are identical using the following command on each file.
[bash]
sha256sum hg38.fa.gz
[end bash]
The result I got for each file was as follows:
[checksum_result]
c1dd87068c254eb53d944f71e51d1311964fce8de24d6fc0effc9c61c01527d4 hg38.fa.gz
[end checksum_result]
Hence, I now have the means to back up the entire (compressed) human genome such that all of its data can be saved to the WayBack Machine in ~5 megabyte increments.
This journal entry file and each of the 5.2 megabyte chunk files which comprise the original hg38.fa.gz file are available in the following karbytes “micro directory”:
[micro_directory_url]
https://karbytesforlifeblog.wordpress.com/raw_github_files_directory_reference_human_genome_fasta_hg38/
[end micro_directory_url]
This web page was last updated on 01_OCTOBER_2025. The content displayed on this web page is licensed as PUBLIC_DOMAIN intellectual property.
KARBYTES_FOR_LIFE_BLOG 