DISQUS

I don't think refseek is that original (especially since it's name is too close to the NCBI reference sequence database (http://www.ncbi.nlm.nih.gov/RefSeq/). The quality of results though is pretty good, especially for casual scientific searches (for "expert" life science searches, I prefer something like NextBio or GoPubMed or PubMed itself).

What I liked most about RefSeek ... speed!!! It's pretty quick for this space, at least when I tried it

Robert,

But wouldn't you say, using your hurricane analogy, that your reaction has been, "we are all going to die", instead of a let's be calm and collected and see what are the steps we need to take to weather out/survive the storm. I was never that interested in behavioral economics. Given the reaction of people to what's been happening, I am now.

Thanks for noticing The Life Scientists. We quite like our FriendFeed community

I am with you. That is NOT a word

I am not sure there was anything new in there. As a product manager, the best PR was getting my customers to talk about my products, and for one simple reason (well two); they liked the product and they valued their relationship with the company as it would help them be successful at their day job. if you can't achieve that, you can hire the best PR people and it won't be worth much

Moments like these are the ones that put our often silly tech world arguments and disputes into perspective.

Michael

Thanks a ton. That's worth a blog post in itself.

I'm looking forward to meeting you at BioIT world as well.

Leo,

I have been waiting for you to do something like this for a long time. Can't wait to see the show and the evolution of online video in your hands.

It's easy enough if you know how Dapper works. The way you set up the Dapp doesn't really change. You just need to make sure the attributes you assign come from an accepted namespace.

Actually that's probably the most important part. Without the collective intelligence, it's probably not going to mean that much.

Decided to leave a video comment about Freebase :)

[viddler_video=49c7b76b]

Twine is (at least not yet) not a developer platform, and is more consumer focussed. Calais on the other hand is a text analytics platform for entity extraction. I am not sure it works for the kind of information you want to pull out, but you should check out out.

Of course, you could always try Freebase as well

They're definitely playing it safe. It's got a ways to go in terms of usability. I'll post my own review once I have built up a sufficient network and content.

The way I look at it, look at it as an application that extracts entity information from your content and uses that to build a graph of linked data (under the hood) and serve up recommendations and content based on that. It reminds me of Calais to a degree, but it's meant to be an application and not a service (at least at this point). Will be very interesting to see how it evolves.

I completely agree. It's something that people don't understand, including some in the life sciences.

Current in silico methods use too many approximations. It's amazing that they do as well as they do. I don't think you can wait for in silico methods at all, but they can be used for candidate selection or to try and understand mechanisms and of course, in 30 years or so when our computers are fast enough then perhaps everything will be designed computationally.

[viddler_video=53d3e67a]

lol ... too funny

I had heard of dailystrength but not mdjunction. As David Hamilton wrote earlier, there are a ton of Health 2.0 companies. Lots of overlap and in time, I believe things will settle down/consolidate.

Trusera is local to Seattle and I got to hear the backstory when the company was first founded. Plus it has quite the rockstar team behind it. IMO, the ones that will do best are the ones that go with a "less is more" approach. Too many sites just try and do too much. In this case, it's all about personal stories with some expert content. Makes a lot of sense to me. In the end success lies in building enough of a user base for the wisdom of crowds to be relevant and in execution.

Depends. I was doing QMMM around protein active sites 10 years ago, but in the docking community with an emphasis on throughput, in general docking functions used a soft potential and a mostly rigid receptor with flexible ligands, and scoring functions were rule or knowlede based. As computers have become more powerful and with an emphasis on quality it's become more common to use flexible receptors and higher order potentials. QM is still rare (too slow). As someone once told me, you have to be faster than the medicinal chemists :).

In this case, it was more of a personal win, to get the structure-based drug design folks to start thinking of using MD-based search schemes and MM-based potentials. The above talk probably uses a modification of Charlie Brook's CDOCKER method and a full workflow (conformer generation, etc) in Pipeline Pilot.

That could be the case, and for some reasons I find the Yahoo properties to be very discombobulated. For whatever reasons, I don't think of Flickr or del.icio.us as Yahoo properties (and in the wake of the Microsoft situation, I really worry). My #1 objection to the Microsoft situation (which I think is going to get done) is that Yahoo is such an open source company.

Maxine

You're quite welcome :)

I didn't really pay attention to the essays, since they were pretty "light". I picked out the items they had selected and wrote some stuff on the ones of interest at this end.

[viddler_video=37a1cc04]

Oh ... and thanks for commenting. Didn't even think anyone would notice :)

Toby

I am still trying to figure out the most effective way. The current thought is to use it as a queryable backend for data from e.g. docking experiments or bioinformatics search results. Essentially as the equivalent of a triple store. Then one could build apps, where the query engine goes straight to a Freebase store.

I am enjoying it already. As for the visual editor, pretty much the first thing I do with any service is turn of the WYSIWYG editor. Markup is so much faster and effective.

The CC and OpenID were key reasons. With exceptions, I am trying to avoid signing on to new services that don't offer support for both.