HNHacker News
TopNewBestAskShowJobs

mfld

762 karma · joined October 25, 2022

submissionscomments
mfld··on Ask HN: Why did Python win?
This is true. Bioinformatics was full of Perl scripts for a variety of (text) analyses. I however remember well that many students began to hate it soon after working with it, as it was very difficult to understand existing code. So, when given a choice, many choose Python as an alternative. And stayed with it.
mfld··on The complete sequence of a human Y chromosome
Yes, there is extra information in the first cells, in particular regulatory elements such as miRNAs. The headline here is epigenetics.
mfld··on The complete sequence of a human Y chromosome
You could download the raw sequence reads (FASTQ files) and map them to the new T2T-Y. Probably it makes more sense to wait for a new release of the T2T human genome reference, as mapping to a subset of the genome increases false positives. It helps to have some genome bioinformatics knowledge to correctly handle those T2T references.
mfld··on Ask HN: What's the state of consumer genomics in 2023?
Regarding the accuracy question, I have a pretty recent article: https://mfasold.net/blog/assessing-nebula-genomics-dtc-wgs/
mfld··on Ask HN: Could you share your personal blog here?
https://mfasold.net/

Just started, after years of not blogging. On bioinformatics, genomics, software.

mfld··on Design and synthesis of a minimal bacterial genome (2016)
A 2022 review on synthetic genomics from the same team is to be found at https://doi.org/10.1016/j.cell.2022.06.046 .
mfld··on OpenWorm – A computational model of C. elegans worm
I like these bottom up approaches, as they demonstrate very well how much we _don't_ know yet about life. Important to mention here Craig Venters minimal cell project syn3.0, where the team synthetically created a livable cell comprising 473 genes. It was done to a large part with trial and error, the function of many of those genes is still not known. A recent review from the same team is to be found at https://doi.org/10.1016/j.cell.2022.06.046 .
mfld··on Tell HN: Nearly all of Evernote’s remaining staff has been laid off
Let's assume they shrink from 300 to 30 employees. After some time they adjust internal processes, focus on the original goals and main use cases, weeding out unnecessary features, and having stable applications for Desktop, web & mobile. 30 people should be sufficient to maintain & support & maybe even improve it a bit for a long time given a healthy revenue stream from loyal customers.

Not sure if something like this ever happened in the history of software. :)

mfld··on Sci-fi author 'writes' 97 AI-generated books in nine months
True. I found the description of his "writing process" somewhat interesting, how he circumnavigates current LLM limitations to create long-form content. Yet, you might be right that the number of 574 copies sold should not be attributed to the actual book content itself, limiting the article's value.
mfld··on Deep sleep may be the best defense against Alzheimer’s
Related question: I have my genome but I find that the reports on polygenetic scores that you receive from [popular consumer WGS sequencing provider] are lacking accuracy. Is there a platform/ community of personal genetics enthusiast that crowdsources such analysis?
mfld··on Mojo – a new programming language for AI developers
Another "Python with high-performance compiled builds" would be https://github.com/exaloop/codon.
mfld··on I don't want streaming music, I just want to stream my music
I store my music on a QNAP NAS. I can listen to it from everywhere either with the provided web app or on my phone with the QMusic mobile app. Both do their job. I like that you have a basic rating/tagging function, which however is stored in an internal database instead of the file metadata. You can also directly delete files.
mfld··on Creator of Catan, Klaus Teuber, has died
Cities & Knights makes the game richer, absolutely recommended for advanced board game players. I heard that this was actually the original game as he designed it, but the publisher requested a more beginner friendly version, leading to the base game as we know it.
mfld··on A ChatGPT Emacs Shell
Another one specifically for org-mode would be org-ai:

https://github.com/rksm/org-ai

mfld··on DNA, AI facial reconstruction, and grit identified Somerton Man 75 years later
Written like this, it doesn't make sense since a SNP has a length of 1. However, you need longer DNA fragments to be able to correctly determine the SNP, and old DNA tends to be fragmented. They are likely referring to those fragment lenghts.
mfld··on Strife at eLife: inside a journal’s quest to upend science publishing
It will also change the process from the perspective of an author. If you are not fully sure about the impact of your publication, you might start submitting it to the journal with the highest possible impact. And, after a rejection, continue submitting to the next best journal, until acceptance. The new process by eLife would hence massively cut the workload on authors, besides editors and reviewers.
mfld··on Asciiquarium.live – An ASCII Aquarium in your terminal without any dependencies

  curl "http://asciiquarium.live?cols=$(tput cols)&rows=$(tput lines)"
mfld··on A cartography of human histology is in the making
One should point out that there were several previous large-scale initiatives to characterize human tissues and cells on a molecular level. For example, GTeX [1] provides gene expression profiles from 54 tissues. However, thanks to single-cell RNA sequencing, it is now possible to look at individual cell types.

[1] https://gtexportal.org/home/

mfld··on How we plan to de-extinct the Dodo bird
I was wondering that, too. But:

> Shapiro said that she had already completed a key first step in the project — fully sequencing the dodo’s genome from ancient DNA — based on genetic material extracted from dodo remains in Denmark.

> The next step was to compare the genetic information with the dodo’s closest bird relatives in the pigeon family — the living Nicobar pigeon, and the extinct Rodrigues solitaire, a giant flightless pigeon that once lived on an island close to Mauritius. It’s a process which would allow them to narrow down which mutations in the genome “make a dodo a dodo,” Shapiro said.

(https://edition.cnn.com/2023/01/31/world/dodo-bring-back-fro...)

From a technology perspective - due to recent advancements in ancient genome sequencing and synthetic biology methods - I find it's a good time to start such a project. Also note that lots of avian genomes have been sequenced in the last years.

mfld··on Ask HN: AI and Eugenics
Note that there is already an application that helps people identify, based on DNA analysis, who they should NOT date (Islendiga-App). However, I cannot imagine that the opposite would provide any practical nor useful application.
mfld··on We're wasting money by only supporting gzip for raw DNA files
If you have re-sequencing data of model species (which applies to >80% of generated sequencing data), the storage issue is often solved using CRAM/BAM formats. The FASTQ can be reconstructed if unmapped reads are stored in the file.

More general (pre-alignment) sequence compression methods never really took of (e.g. https://github.com/BEETL/BEETL). Probably because it helps so much to have common format that most workflows can start with. Here, the replacement of gzip with zstd would be a lower hanging fruit.

mfld··on We're wasting money by only supporting gzip for raw DNA files
The best fit is probaly the Oxford Nanopore MinION. However, it would still be good to have a range of lab equipment available.
mfld··on What’s Left in the Apple Silicon Transition
A while back I built a linux box for bioinformatics and was happy with the performance [1]. You might need to look for suitable main boards to support your RAM requirements, though.

[1] https://www.ecseq.com/blog/2020/Whole-Exome-Analysis-on-a-Ma...

mfld··on What software that you use daily “sparks joy” for you?
Emacs - getting better since 40+ years.

Ripgrep - Fast & exactly the results I need.

Google Chat (!) - simple and just works.

mfld··on Ngrok Raises $50M
Wow. I am a customer myself and like the service. Yet, I am surprised by the size of this investment, as it doesn't look like a venture case to me. Wishing Alan and his team the best.
mfld··on Ngrok Raises $50M
Maybe you can find something on that list: https://github.com/anderspitman/awesome-tunneling
mfld··on Ngrok Raises $50M
I can confirm this. We work a lot with hospitals. When we want to let them try our services behind ngrok domains, they are often blocked. Unfortunately.
mfld··on Ask HN: Good Bioinformatics Books?
If you are coming from the computer science angle, I can recommend "Computational Molecular Biology: An Algorithmic Approach" by Pavel Pevzner.
mfld··on Consider working on genomics
I have a practical example to explain what, at first glance, doesn't seem to add up. Many genomics analysis contain a step to clean up DNA sequences. For this "Read Trimming" step there exist no fewer than 40 open source tools. Let's say you need this step in your project and use one of these tools. You find an issue the original creator is unwilling to fix: just choose another tool, problem solved. You find that the tool does not perform well enough on your data: choose another tool, problem solved. So while the articles points are true, they in practice often don't lead to a real pain.
mfld··on Consider working on genomics
Great post - which contains the answer to many of the questions raised in this thread. I am working in this area as well. There is "next to zero demand for tool development" because there are great open source tools. Only in rare cases (e.g. Illumina Dragen), a commercial software adds significant value that the audience is willing to pay for.
← PreviousPage 5 of 6Next →