Here is a round up of some of the stories on the Einstein robot that learned to make its own facial expressions, thanks to a mirror, machine learning, and some serious UC San Diego brain power.
LiveScience.com
http://www.livescience.com/technology/090710-robot-expression.html
Popular Science
http://www.popsci.com/scitech/article/2009-07/einstein-robot-teachest-itself-smile
TG Daily
http://www.tgdaily.com/content/view/43157/135/
Wired
http://www.wired.com/wiredscience/2009/07/robotsmile/
Technology Review blog Kristina Grifantini (DK) (B,D,E)
http://www.technologyreview.com/blog/editors/23825/?nlid=2173
Technology Review video stream: (DK) (B,D,E)
http://www.technologyreview.com/video/?vid=388
Tehran Times
http://www.tehrantimes.com/index_View.asp?code=198622
KPBS radio
http://www.kpbs.org/news/2009/jul/10/ucsd-robot-learns-human-facial-expressions/
CNET Gadget blog: “crave” Einstein bot: E = mc smile
http://news.cnet.com/8301-17938_105-10286624-1.html
DViCE
http://dvice.com/archives/2009/07/a-robot-that-ca.php
GIZMODO
http://www.gizmodo.com.au/2009/07/robot-einstein-still-learning-to-smile/
gizmag
http://www.gizmag.com/robot-learning-facial-expressions/12218/
Switched
http://www.switched.com/2009/07/12/self-taught-einstein-bot-learns-to-smile-creepily/
MSN 9 News
http://news.ninemsn.com.au/glance/836366/robot-einstein-learns-to-smile
Popular Science video stream
http://www.popsci.com/scitech/article/2009-07/einstein-robot-teachest-itself-smile
Engadget
http://www.engadget.com/2009/07/10/einstein-robot-learns-to-smile-teaches-us-how-to-feel/
The Einstein robot video on YouTube had more than 73,000 hits as of Friday July 17
http://www.youtube.com/watch?v=UBUtxfUY_w0&eurl=http%3A%2F%2Fwww%2Egizmodo%2Ecom%2Eau%2F2009%2F07%2Frobot%2Deinstein%2Dstill%2Dlearning%2Dto%2Dsmile%2F&feature=player_embedded
Friday, July 17, 2009
Thursday, July 16, 2009
New Drugs from Natural Compounds
Congratulations to Julio Ng (front) and Nuno Bandeira (back), the UC San Diego researchers and co-lead authors on a Nature Methods paper describing computational and experimental advances that enable researchers to quickly and inexpensively determine whether natural compounds collected in oceans and forests are new -- or if these pharmaceutically promising compounds have already been described and are therefore not patentable.The UC San Diego web-based tools for sequencing nonribosomal peptides (at no cost to researchers at nonprofit organizations) are available at: bix.ucsd.edu/nrp
This research appeared this week on a number of media outlets, including:
Scientific American (story by Brendan Borrell)
http://www.scientificamerican.com/article.cfm?id=new-algorithm-antibiotic-hunt
Genetic Engineering News
ttp://www.genengnews.com/news/bnitem.aspx?name=58159197
Natural Products Industry Insider http://www.naturalproductsinsider.com/news/2009/07/new-drugs-faster-from-natural-compounds.aspx
Genome Web Daily News
http://www.genomeweb.com/proteomics/ucsd-team-develops-de-novo-sequencing-dereplication-method-decipher-nonribosomal
Wednesday, July 15, 2009
NanoEngineering for Fighting Cancer Video Wins!!!
The University of California, San Diego NanoTumor Center and NanoTecNexus (NTN) (formerly NanoBioNexus)—a leading nanotech education organization—won the 2009 Bronze Telly Award for the production of a video on approaches to fighting cancer using nanotechnology. Read the story here.
The three minute video, entitled “Fighting Cancer with Nanotechnology,” is embedded below and can be viewed at NanoTecNexus, YouTube and around the Web. The video was produced in collaboration with Mindeliver Media, a San Diego-based media company specializing in pharmaceutical, biotechnology, and science industries.
Tuesday, July 14, 2009
A Semiconductor Quandary
UC San Diego scientists including Jacobs School research scientist Mark Tillack (above) are using laser plasma-produced light sources to explore performance improvements of critical inspection tools for the semiconductor industry, which ultimately will enable industry to pursue even better and faster chips. While optical lithography is being pushed to its limits in the semiconductor industry, there is a growing concern whether metrology tools can keep pace for creating and inspecting the new generation of devices. Read more here.
Monday, July 13, 2009
Science paper "extended introduction" to new Nature Methods paper on Drug Discovery from UC San Diego researchers
If you want to have a broader understanding of the importance of the computational breakthrough described in a Nature Methods paper published today by UC San Diego researchers, then take a look at a paper published last friday in Science magazine.
The Science Review highlighted the fact that the number of drugs from natural products is dropping even though there are vast numbers of untapped natural compounds that could lead to new drugs.
Below I have pasted a few reactions and parenthetical comments on this Science paper from one of the co-lead authors on the Nature Methods paper: Julio Ng, a doctoral student in Bioinformatics at UC San Diego.
Also, there is a cool underwater drug discovery photo of one of William Gerwick, one of the Nature Methods authors, here:
http://cse-ece-ucsd.blogspot.com/2009/07/algorithms-answer-hot-or-not-question.html
The link to the original press release is here:
http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=866
“The Science paper is very pertinent to our work. My impression is that the Science paper is an extended introduction of our paper, explaining the motivation of our work. Of course, we only attempt to automate the characterization of a set of natural products with a very specific structure (cyclic peptides). In the Science paper, they also discuss the potential of polyketides for commercial drugs. However, a main point of the paper is that drugs based on natural products seem to be declining in recent years. One of the reasons for this decline is that the model of natural product screening is not amenable to high-throughput screening (HTS). This is where we want to bring our software to help the high-throughput identification and dereplication of natural products. I really like the concluding sentence where the authors state that it is not that natural products are a dead end from drug discovery, far from it, they are a vast resource that remains unexplored. And I think we would like to provide the means to help achieve this end.”
-Julio Ng, a co-lead author on the Nature Methods paper and a doctoral student in Bioinformatics at UC San Diego.
Some relevant quotes from the Science paper with added comments from Julio Ng, a co-lead author on the Nature Methods paper and a doctoral student in Bioinformatics at UC San Diego.
“A prevailing sentiment in many pharmaceutical organizations is that screening of natural product sources is a difficult effort with a high probability of duplication; that is, the result may be a known compound that cannot be patented.” (This is the problem of dereplication – Julio Ng) [This is Daniel Kane typing now...the "problem of dreplicatioin" is one that the UC San Diego researchers tackled in their Nature Methods paper].
“Clearly the biological resource is there, but access and examination are problematic, especially if there is pressure for a short time frame for discovery of new leads.” (I think mass spectrometry, coupled with computational tools can speed up this process significantly. Mass spectrometry itself has advantages of being high-throughput, require small quantities of the sample, the data is amenable to computational tools for analysis, etc –Julio Ng)
Below I have pasted a few reactions and parenthetical comments on this Science paper from one of the co-lead authors on the Nature Methods paper: Julio Ng, a doctoral student in Bioinformatics at UC San Diego.
Also, there is a cool underwater drug discovery photo of one of William Gerwick, one of the Nature Methods authors, here:
http://cse-ece-ucsd.blogspot.com/2009/07/algorithms-answer-hot-or-not-question.html
The link to the original press release is here:
http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=866
“The Science paper is very pertinent to our work. My impression is that the Science paper is an extended introduction of our paper, explaining the motivation of our work. Of course, we only attempt to automate the characterization of a set of natural products with a very specific structure (cyclic peptides). In the Science paper, they also discuss the potential of polyketides for commercial drugs. However, a main point of the paper is that drugs based on natural products seem to be declining in recent years. One of the reasons for this decline is that the model of natural product screening is not amenable to high-throughput screening (HTS). This is where we want to bring our software to help the high-throughput identification and dereplication of natural products. I really like the concluding sentence where the authors state that it is not that natural products are a dead end from drug discovery, far from it, they are a vast resource that remains unexplored. And I think we would like to provide the means to help achieve this end.”
-Julio Ng, a co-lead author on the Nature Methods paper and a doctoral student in Bioinformatics at UC San Diego.
Some relevant quotes from the Science paper with added comments from Julio Ng, a co-lead author on the Nature Methods paper and a doctoral student in Bioinformatics at UC San Diego.
“A prevailing sentiment in many pharmaceutical organizations is that screening of natural product sources is a difficult effort with a high probability of duplication; that is, the result may be a known compound that cannot be patented.” (This is the problem of dereplication – Julio Ng) [This is Daniel Kane typing now...the "problem of dreplicatioin" is one that the UC San Diego researchers tackled in their Nature Methods paper].
“Clearly the biological resource is there, but access and examination are problematic, especially if there is pressure for a short time frame for discovery of new leads.” (I think mass spectrometry, coupled with computational tools can speed up this process significantly. Mass spectrometry itself has advantages of being high-throughput, require small quantities of the sample, the data is amenable to computational tools for analysis, etc –Julio Ng)
Algorithms answer "Hot or Not?" question for natural compounds
Hot or Not? New or Not? Discovery or Not? Hot Pharmaceutical Prospect or Not?
A researcher isolates a natural compound with promising antimicrobial properties from ocean water. But is it a discovery? Or has the compound already been described and patented? University of California, San Diego researchers have invented computational tools that enable researchers to rapidly and economically answer the ‘is it new or not?’ question for promising drug targets.
Scientists will finally be able to rapidly identify and characterize ring-shaped nonribosomal peptides (NRPs)—a class of natural compounds of intense interest due to their potential to yield or inspire new pharmaceuticals. This advance should speed the discovery of natural compounds produced by organisms such as blue-green algae that could lead to new drugs.
This new work will be published online on July 13 in the journal Nature Methods.
Full press release is on the Jacobs School of Engineering web site:
http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=866
A researcher isolates a natural compound with promising antimicrobial properties from ocean water. But is it a discovery? Or has the compound already been described and patented? University of California, San Diego researchers have invented computational tools that enable researchers to rapidly and economically answer the ‘is it new or not?’ question for promising drug targets.
Scientists will finally be able to rapidly identify and characterize ring-shaped nonribosomal peptides (NRPs)—a class of natural compounds of intense interest due to their potential to yield or inspire new pharmaceuticals. This advance should speed the discovery of natural compounds produced by organisms such as blue-green algae that could lead to new drugs.
This new work will be published online on July 13 in the journal Nature Methods.
Full press release is on the Jacobs School of Engineering web site:
http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=866
More photos:
caption for above photo:
Nuno Bandeira (left) and Julio Ng (right) are UC San Diego researchers and co-lead authors on a Nature Methods paper describing computational and experimental advances that enable researchers to quickly and inexpensively determine whether natural compounds collected in oceans and forests are new -- or if these pharmaceutically promising compounds have already been described and are therefore not patentable.
********************************************************************************
credit for photo above:Scripps Institution of Oceanography, UC San Diego/ Cameron Coates
caption for photo above:
William Gerwick collects natural compounds in search of new pharmaceuticals. Gerwick is an author on the new Nature Methods paper and a professor with the UC San Diego Scripps Institution of Oceanography Center for Marine Biotechnology and Biomedicine and the UCSD Skaggs School of Pharmacy and Pharmaceutical Sciences. Read about Gerwick’s work to discover drugs and protect Panama’s natural and cultural resources at:
http://explorations.ucsd.edu/Features/2009/Discovering_Diversity/images/07_2009_Feature.pdf
Thursday, July 9, 2009
Robot Learns to Smile and Frown: More Images!
more photos from the robot facial expression story I posted yesterday:credit: UC San Diego / Erik Jepsen
“As far as we know, no other research group has used machine learning to teach a robot to make realistic facial expressions,” said Tingfan Wu, the computer science Ph.D. student from the UC San Diego Jacobs School of Engineering who presented this advance on June 6 at the IEEE International Conference on Development and Learning.
Download the paper at:
Subscribe to:
Posts (Atom)