The annual Society for Neuroscience Conference begins in a week! I present my newest work at the final session on the final day – not sure if anyone will stick around 
Poster 889.08 – Epigenetic mechanisms in temporal lobe epilepsy
Wed Afternoon session 11/16
Come by and say hello.
Yeah, we got the cover! Granted this particular journal doesn’t have very large images on the cover and they only have a very miniature version of the cover online (so I had to scan it and and put it on my blog for anyone to see it). BUT still, my pretty pictures made the cover
Oh yes, and the research article is good too (p. 454, if you’re into that kind of thing).
Here’s a HUGE version of the graphic strip that you can click on for way too much resolution. This was a very early graphic that I made, I feel like I have made much better pictures since then, but this is basic and to the point of what we saw in our research.
So many tablets coming out!
They are getting better, but they’re not really all that different than the tablets of decades past – However, they certainly are gaining consumer support. I mean, I got one of the $99 HP touchpads in the fire sale. For that price why not. And heck I love the thing (I’ve played with the iPad and most of the others, they’re all basically the same). But I can’t help thinking the bottom might fall out of these tablets – they aren’t incredibly useful until they replace laptops. On the couch, when I can reach for either a tablet or laptop, I still reach for the laptop alot out of ease of use (which is supposed to be the tablet’s selling point).
No doubt, some day they WILL replace laptops. I was really hoping for the brain interface thing (LINK TO POST) – that really would have pushed tablets along, obviating the keyboard a bit. The Samsung Series 7 Slate pictured above includes a pressure sensitive Wacom digitizer, which may help a bit…but it’s same old/same old (but we’ll see how consumers like it, maybe if Apple ‘invents’ the digitizer for the iPad then it’ll take off).
So it’s been awhile, let me revise my ELN dream!! I’m debating keeping the brain interface in there because it makes so much possible… we’ll see, this is a dream after all….
ELN software
Interestingly, as I’ve used ELN software over the years (and, in the past, I wouldn’t have really imagined this is the view I would have), the ELN software is not very important at all. What!? So (as I look for a good job with this PhD I got), I may not be winning any jobs from ELN companies by saying this (which I would love to have by the way, if you’re reading). But I’ve tried alot of them out, I’ve used with my project members mostly Onenote, and have organized many network shares for other labs – no real ELN software at all. All this, and there isn’t a big difference. Now searchability and accessibility… and real-time accessibility are what make a huge difference – and these things have actually progressed quite well. Great software is a plus and makes the whole idea of an electronic lab notebook more palatable, but so many other aspects are more important.
Collaboration, real time access
So the greatest thing about an ELN is to be able to actually be sharing the lab notebook over time and space, and to be able to search and find data quickly and easily. THE CLOUD is being shoved down our throats, but (although clouds usually block light) THE CLOUD is a bright spot for the future of ELNs. I get it though, this IS software nowadays – the cloud and apps, so disregard my above derogatory opinion on ELN software (and give me a job). I’ve collaborated with people all over the would on research projects, using skype/video calling to collaborate – sometimes sending/receiving files (with some daring folks actually having a revision of the paper in the cloud!). But I’m sure the future lies in a Facebook style ELN, where they can see all the data and protocols from our online lab notebook, can see if we’re there to chat or video chat about things, make notes, send messages, put forth ideas, quash ideas That’s the value prospect for ELNs in the future, the software of organizing and recording the data – you know, that’s not much different than the software for organizing and recording non ELN type data, might as well just implement and use the heavily supported and developed products the rest of the world uses.
Data recording Hardware
So this is where the explosion of smartphones and tablets in the consumer space has really moved things along…finally. But lets go beyond that, the cloud and wireless connectivity offer us many possibilities. Lets forget the brain interface 
and even tablets. How about science goggles. Built in camera/sound recording, able to project a personal screen of your lab notebook to view, maybe touch sensitive controls on the side (although maybe we don’t want to be touching near our face all the time – voice activated maybe…brain interface, no, no), maybe a Kinect type interface (write things or make gestures on any surface and the camera watches you), … all connecting wirelessly to the cloud ELN real-time. Maybe someone in France sees online that your doing some experiment and says ‘can I watch?’ or discusses some alternatives to the protocol. Crazy, but not far fetched at this point. Well, at least in the very near future the tablets may give ELN productivity a boost.
Outside the box – new ideas
Well the cloud and science goggle idea is ‘boxless’, but let’s really push it. What would be really cool and crazy ELN-wise in the future.
AI – Artificial intelligence I’ve been on a programming kick lately (as I was looking at a software development job that fell through), and I spent some time learning about AI programming. Maybe a real-time AI checking your protocols and results and times, could chime in and suggest ‘are you sure you didn’t mean to use this sample’, or ‘based on previous experiments, maybe a more stringent wash would be more successful’. Heck there’s a computer making it’s own discoveries already (LINK TO STORY). Why not let the computers HELP us instead of having them take over the human race apocalypse style. Plus the more cloud and connected this becomes, the more likely it would suggest things based on other people’s work – ‘so and so added more NP-40 to this wash solution and it worked like a charm’. Actually getting a simple AI to just provide you with related info at opportune times would be very useful and not that difficult to implement.
Real-time Inventory/equipment control – I have worked in many different labs and the initial learning curve isn’t hampered as much by learning a new technique (heck that’s fun), but learning where everything is! Even in my current lab of over 4 years, sometimes it’s hard to find what I’m looking for. With correct hardware/software/ELN integration hopefully all this could really reduce a blackhole of productivity – and considering how the scientific community loves to have short term post doc positions, graduate positions, and rotations, empowering new (and old) lab members in this way could prove quite beneficial.
Patent/Drug Licensing Social network – What? This is more a policy change that needs to occur that can then be implemented using ELNs of the future. Companies/Scientists need to recoup research expenses – patents are NOT bad, despite what the media may want you to believe. However, there are flaws, such as a new drug that works great will be found and nobody does research on it after its approved for decades. Another problem is, the drug is expensive (for recouping) to buy, but inexpensive to make usually – and 3rd world countries can’t benefit from them during this time. The idea of licensing the drug to third world countries (with the requirement the drugs not LEAVE said country) doesn’t hurt anyone and actually big pharma makes a little extra. But how about the research? How about the companies farm out availability of there drugs under licenses, to researchers checking different diseases/models under agreed upon rules (like no competing product, no publication,…) – this could help fund struggling researchers, but also help us to understand the science behind some of these drugs (leading to better drugs). Anyway, I’m off target a bit with regards to ELNs, but come on, this is the future all the technical and interfacing problems are resolved The thing holding back discovery in the future is how to make discovery lucrative and not hinder advances that should compound upon other advances – as well as be responsible about ‘sharing’ the cures we find with other countries.
Okay, I don’t want to tire my few readers , so I’ll save more for later. Leave comments with your ideas and critiques!
Yay, I presented my dissertation yesterday (see below). Now I’m a doctor
Added 08/18/2011: Got 45 minutes? Here’s the video
link
Here’s my sketchflow quick mock up for a windows phone app (game). Click here.
The idea being direct competition between people for these brain teaser type games and quizzes. Fun and good brain exercise.
I’ll develop it if I win a developer phone
I’ve been running my current lab research project for about 4 years. I started off with a small group and was sharing Onenote 2007 from one computer to two others on our network. I tried a couple times to synch to different cloud services (tried Groove a LONG time ago), but it always seemed easier (and better) to run from a shared computer. I have a lot of links to files I want to open in other apps like (Prism (statistical software), Photoshop, excel,…) Making sure those links work on a shared computer was sometimes tricky, but making sure they work from the cloud was even more difficult. Not impossible, but difficult to the point that training people to do it was not worth the time.
Things have come along somewhat since I started. The linking is still not great (it would need to be embedded to really be ‘easy’, which I would love – they did that a little with the newer PowerPoint), but wow, after playing with it awhile I LOVE the new abilities and were I not graduating and moving on in a month, I would move everything to SkyDrive.
Gladinet!
Linking is just as easy (if not easier… actually, now that I think about it, it IS easier) using Gladinet. Every shared computer can just map the SkyDrive folder containing the LabNotebook to a local drive and now when a link calls the the folder it is the same call on every computer, it works GREAT!
A couple huge drawbacks-
Migrating my current 1000+ pages notebook would be VERY difficult – fixing links, etc.. One reason it would be difficult is that to just copy the OneNote notebook to the SkyDrive doesn’t give you access to it via the Web app!? Strange, but I think this may be a side effect of the – we’re phasing out users ability to see/know file extensions. For once a file is in the cloud you can’t change its extension in the cloud (you can change it from your Gladinet mapped drive and synch it though). But still, it recognizes your copied OneNote section as a OneNote section, you just cannot open it in the webapp. To migrate you can make a new OneNote book, open it into your local OneNote application, and then copy sections over to your SkyDrive OneNote notebook.
SEARCH!..in the web app. For the love of… This is the whole reason I have an electronic lab notebook! To be able to find stuff quick (also to make collaborations easier,..blah, blah). Anyway, I know the idea is probably to give people some reason to actually buy Office, since the web apps are pretty functional. But come on, I own it and search is huge. Granted, I can always press the little ‘open in OneNote’ button on the webpage and then search in OneNote.
Anyway – there are positives as well – even above and beyond the normal ELN positives I’ve always listed.
THE WEB APP!! It could be better, but give me a break, this is huge. I can show ANYONE my ELN, from anywhere with this. It’s not some nebulous idea that people say to ‘hey, that does sound like a good idea’. I can show them, they can edit, some collaborators in France recently saw it and replied ‘you mean, I can just look up what you did, how much drug you delivered on each experiment, whenever I want?’ I said ‘NO, I mean you can look that up and do your experiment, record it in here as well, and we can (almost) real time see how these experiments are working’! J **much easier to convert the technophobe or reluctant user
…the web app’s ‘Show Author’ The show author option showed up (in a useful manner) in the upgrade from OneNote 2007 to 2010. But still it would require a logout/login for each individual user so it turned into more of a ‘which computer was used to input this info’ (which is still useful). But more useful is the fact that, a student or lab tech can just log on to their individual account, access the web app and any input has their actual name (see leading picture). Still the way I would use it (for the most simplicity) I would probably use Gladinet and the local OneNote program so it will still be recording just the computer (since there’s no way we’re getting people to login and logout all the time). Show’s some potential though!
…I’m tempted to migrate everything anyway J, just so I can show off our work direct from the webapp, then open a PowerPoint presentation direct from the actual notebook, then go into our normal presentation of the work. But then some people get offended by my over the top geekery. I try to keep them focused on the actual research work when I’m presenting L
Big year for my family! We’re finishing up here at UC Irvine with my PhD, looking for a new job and possibly moving (contact me if you hear of something), and now the family has grown from 4 to 5! Nathan was born June 1st. 3 boys in the family! (Well 4 if you count me). It’s great fun (and sometimes great drama).
update:
HAHA! Got it!
With some Mathematica help on solving summations. I managed to see the ‘easy’ pattern everyone talks about in the forum that gets unlocked upon solving it! Boy I really over did this one
(4 a^2 (2 + b) + a (b (9 + 5 b – 5 c) – 8 c) -
3 b (2 + b) c + (4 a – 3 c) (-2 + a – Mod[b, a]) Mod[b, a])/***
That’s my equation (simplified believe it or not). I took out a piece as to not give away the answer.
No recursion, no anything, executes immediately.
I’ve been practicing different programming languages by solving problems on this Project Euler site. A new problem just went up, so I thought I would show my stuff and be one of the first to answer it
Well, I figured out much of the clever stuff I think and actually solved to 9 -digits I swear! 365911736, by summing the ‘reduced’ crazy function up to b=083284007 the last 9 digits of 7^21…Actually I went to the last 11 digits to make sure 365911736 was staying steady.
But my answer’s wrong (or I wouldn’t be posting it here, because they don’t want people spoiling things). Everything checks out though? Weird, I’ll have to come back to it later with fresh eyes.
Again thwarted!!
Here’s my crazy code, not cleaned up. I figured out how to get out of the overflowing recursion of the crazy function, but left in the original so you could see how things evolved.
import time
startt = time.time()
def funa(n):
—-global a,b,c
—-if n > b:
——–return n-c
—-else:
——–return funb(a+funb(a+funb(a+funb(a+n))))
def funb(n):
—-global a,b,c
—-if n%a ==0:
——–return ((((((b-n)/a) + 1)*3)+1)*(a-c))+b
—-else:
——–return ((((((b-n)/a) + 1)*3)+1)*(a-c))+b – (a-(n%a))
**edit: Duh, was able to get rid of the if/else by changing the last part of the formula from (a-(n%a)) to ((b-n)%a)
a,b,c = 21**7,7**21,12**7
#a,b,c = 50,2000,40
total=0
count=0
y=83284007
y=[7,4007,84007,284007,3284007,83284007,4083284007,64083284007, 864083284007,13284007,23284007,33284007,283284007,183284007,383284007,483284007]
x=1
while count —-#for n in xrange(b+1):
—-#print n, funb(n)
—-total+=funb(count)
—-count+=1
—-if count in y:
——–print str(total)[-9:],count
print total
endt = time.time()
print ‘\nRuntime: %dm’ % int((endt-startt)/60), ‘%ds’ % int((endt-startt)%60)
Oh man! I had it. If only I was as fast as these guys to the left!
I got short and long for the 1st problem (but almost everyone did).
I was so close to finishing the second one, (I would have had to get the long set as well at that point in the competition to make the top 1000).
Don’t know if I want to try at 2am my time the round 1C…. Who am I kidding, I refuse to fail. I will do it and I will make the top 1000! I’ll need some hot dogs to give me energy though.
For the Hot dog problem I just needed to figure out how to handle the max vendors per point neighbors… which I figured with 5 minutes left. Couldn’t quite make it.
EDIT – Added
I don’t feel too bad, I did the same as Weeble – and linguo didn’t get the hot dog large set (not quite making the cut). Both are REAL coders (linguo has made it to the Finals in the past, granted maybe he’s not trying too hard this time), so I feel a little better. Not much though
Plus I came home and built this with my kids. Come on, I have some skills… right?! Well, at least my kids think so. Better if they thought I have skills, and I had a programming job, though
Hmmm… Google code jam really took it to me in Round 1A. I thought I would be clever and use strategy to get into the top 1000.
I knew my strengths and weaknesses.
Weakness- only a month of programming experience and in Python (which doesn’t execute as fast as C++), so speed is not my friend
Strength- problem solving ability
So, after practicing some old GCJ problems and surveying the difficulty of past round 1 questions, as well as checking out what scores would get me in the top thousand, I settled on the strategy of just going for the hardest problem. I figured I could somewhat easily get the first problem and probably the second – but in the past you would need to do that (get one and two, or one) moderately fast to get in the top thousand. And speed probably isn’t my strength yet, so I went right to Pseudominion.
BIG MISTAKE.
After reading the problem, it lulled me into believing I could certainly answer it – so I went to work on it. I realized a little too late all the caveats to the planned algorithm I came up with, and then had a heck of a time getting all the extra cases taken care of.
RAN OUT OF TIME. It really got me good. So 0 points!!
The questions this year seem a step harder than past problems in equivalent rounds that I’ve practiced with.
Perhaps I will start with the middle problem in Round 1B (or Round 1C depending on what my pregnant wife and two kids have in store for me tomorrow).
Written by 
Recent Comments