itodd

But it was done in the open. How can it be used as leverage?

I do. I remember posting this to the mailing list. I don't have an example calculating enrichments though. We simply group by read and divide using the frequencies and then plot one enrichment vs another. This way we can see how one sequence enriches between conditions. There's more to it than that but this will produce a plot similar to the one in the attachment in the thread linked below.

https://groups.google.com/a/continuum.io/forum/m/#!msg/bokeh...

Edit: here's a link to the plot I refer to: https://10826817673355204906.googlegroups.com/attach/e6e58ad...

I've used datashader for plotting NGS (Next Generation Sequencing) enrichments. At the time I had to hack together the ability to use the polygon select tool on the data, but it worked and blew my mind.

Very elegant solution to a difficult problem (overplotting).

Just want to say thanks for handsontable. Thanks. It is a great way for my users (scientists who live in Excel) to feel comfortable with the applications we make to help them discover new therapeutics more efficiently.

This is cool and answered a recent itch. I have a pocket operator (po-32) which you can update to play new sounds by using a data over audio protocol. It's really neat. Thanks for sharing.

Adimab | Web Front End Developer (intern/full time) | Lebanon, NH | Onsite

Adimab discovers antibodies. This position is part of the computational biology team. You would be responsible for standardizing interfaces to computational biology tools. No bioinformatics experience required.

I'm not so sure anyone should follow this considering the recent shens over bitfinex and tether.

Tether is supposedly tied directly to USD. Their website says as much and claims transparency through audits. There hasn't been a single audit that has been independently verified.

Right now there have been nearly [800m tethers minted](http://omnichest.info/lookupsp.aspx?sp=31). They are minting them at an alarming rate as seen [here](http://omnichest.info/lookupadd.aspx?address=3MbYQMMmSkC3AgW...).

This is of course circumstantial. I just thought I'd share as I'm stepping back from the markets as a result. @bitfinexed on Twitter seems to be the loudest voice on this topic.

I wrote a bot to do this a few years ago. It would grab the order books from a set of exchanges and execute on the overlap. It had support for maker/taker fees as well as deposit/withdrawal fees for rebalancing accounts. I avoided USD trades as it required so much time to transfer. It worked well and it was a fun project to learn how markets functioned. The risk of missing the trade and the risk of hacks ultimately caused me to stop the experiment

I had BTC in bter when it was hacked. That more than wiped out all of the profits I had made arbitraging.

Thanks for this. I've been running separate instances of jupyter on different ports. This is more elegant.

I'm unable to see the charts:

    Refused to display 'https://plot.ly/~mthwsjc/69.embed' in a frame because it set 'X-Frame-Options' to 'sameorigin'.

Correct.

I recently left my macbook with the TSA after a frantic sprint through security. I was stuck with a windows 7 PC for the week.

Armed with Putty and a browser, I only missed the touchpad of my macbook. I do most of my work in Jupyter and in vim. It's rather liberating to realize you are not tied to one ecosystem. My point is I thought I would never be productive without my mac but that simply wasn't true.

I suppose I am an edge case though as most of my work is performed on my headless linux workstation.

For those wondering, TSA held on to my macbook which I retrieved over the weekend.

I am not an audio engineer but I own a pair of bluetooth headphones and 2 devices that I use them with and it's a nightmare as both devices fight over it. It seems to be the USB problem all over but an order of magnitude worse. I can't just flip the cable over, I need to disable bluetooth temporarily on the device that I don't want to pair.

You also have to pay Equifax $10 to do this. Insane, right?

In my country, it was the winning presidential candidate complaining about fraud.

Good talk from pycon 2017 on hacking the CAN:

https://www.youtube.com/watch?v=3bZNhMcv4Y8

I filled it out. Good luck to you. I hope you'll share your findings.

I just upgraded to a RX480. From an R9 390. Not much gain, but wanted to support the company that is committed to being open.

Me. I'm in the market for a new wireless router. My N66U drops out hourly. It never had great coverage. I have tried different firmware in an effort to save it but I believe it's dead.

Radiolab [as usual] did a great episode on this theory called Cellmates.

http://www.radiolab.org/story/cellmates/

It still seems like a lot (to me).

I did mean hamming. I group by length as insertions and deletions are not expected. I convert to a one-hot vector so I can xor to get the differences.

Thanks for the suggestions. I hadn't thought about splitting the the strings into 3 roughly equal sizes though, that would definitely help narrow the search pool. FWIW, the strings use the protein alphabet (20 values instead of 4).

They are between about 5 and 32 characters. I have converted them to bitstrings (one-hot encoded). They are grouped by length so they are fixed. If you have more information on the algorithm you describe, I would love to hear it. There are not many duplicates in a 1e9 sampling. This is a protein library with a huge diversity for what it's worth.

I'm currently trying to find similar strings (max hamming of 2) between two very large (1e9) sets. Approximate Nearest Neighbor search makes this just barely manageable. Annoy[0] from Spotify deserves a look for anyone who is outgrowing traditional nearest neighbor algorithms.

[0] https://github.com/spotify/annoy

I recently built a kitchen and made the huge mistake if buying appliances from Sears. I am in awe of how badly this company performed.

First delivery attempt, they called the night before to give me a window. Delivery driver called me to make sure I will be home and that he would be there in 30 minutes. They never showed up. After calling I learned the products were not in stock.

They did the exact same thing on the second delivery attempt.

Third delivery attempt they deliver a damaged fridge.

Fourth attempt, they opened the truck and they did not strap down the fridge. Parts everywhere.

Fifth, they got it right. Normally I would have cancelled the order but the experience became a curiosity. How badly could a company mess up? Turns out, a lot.

There was a decent episode of RadioLab on CRISPR not too long ago. Listen here: http://www.radiolab.org/story/antibodies-part-1-crispr/

Sadly the technology is wrapped up in many lawsuits.

It happens. I wrote an arbitrage script that makes a little bit of money every day (only requires me to rebalance the wallets every so often). Only risk is holding whatever is presenting the most lucrative arbitrage opportunity.

You are not fucked. Just breathe. Think about it this way, It's likely that this start up will not be around for long if they make such poor decisions with their most important asset. No matter how you think about it, a very poor decision was made.

You have a masters in CS and 6 years of management experience. You are so far from fucked.

This is fantastic. I have encountered markov chains in my career and always thought of them as a black box. This simple visualization makes it so easy to understand what has previously been so hard for me. Thank you.

Is there an API?