Mobile Lab Management System

LaBLog on a tablet in the lab (Photo credit: cameronneylon)

When I first started working in +Roderic D. M. Page lab, I was fortunate that +Vincent Smith had been looking after the frozen collection of lice. He had inventoried and stored all the different specimens in carefully labelled tubes and boxes. The collection had been carefully built up and catalogued by a number of previous members in Rod's lab in a database called Lousebase. I was lucky because it made it very easy to find the samples I wanted to work on.
Lousebase went through a number of versions, originating from multiple excel spreadsheets used independently by the group, then it was compiled into an Access database, then it made it onto the web via ColdFusion with an underlying SQL database. When ColdFusion stopped playing nice, I made an attempt to write some PHP to interface with the SQL database but finally transferred everything into GoogleDocs here.

Working in the molecular lab which was two flights of stairs up from the office, certainly kept me fit but I dreamt of the day when I could interface with the database via a mobile device, so I put forward a project for an MSc Computer Science student to build such an app to interface with LouseBase on GoogleDocs. Here is a presentation I gave explaining what Nachiket Khandetod did for his MSc project.

Mobile lab app from josephhughes

The main difficulties he encountered was dealing with the Google API and having multiple user accounts accessing the same spreadsheet. At the end of the project, he said that it would have been much easier to link directly to an SQL database. However, he still managed to produce an Android app that updated a Google spreadsheet and that multiple users could work on simultaneously as long as they used a single Google account. He also built data caching into the app so if someone was out of WiFi range, they could store the data to synchronize later.

All the code for the app is open source and available on github but it is not particularly well annotated. One day, when I have time, I will revisit this to build a mobile lab management system the way I think it should be done.

Note:
Unfortunately, the spreadsheet name and Google Account were hard coded into the app, for this reason you will need to change the email (xxx@xxx) and password (xxx) in the following files before creating the .apk:
src/com/ei/auth/BasicAuthenticatorImpl.java
src/com/ei/auth/Account.java

Now that tablets are cheap and common place, it will not be long until they change the way we work, probably to a greater same extent than PCs have.

mini tablet comparison: Google versus Amazon versus Apple

Nexus 7 Home (Photo credit: Stmpjmpr)

I've just done a little table to compare the Nexus 7, Kindle Fire and iPad mini which I thought I would share. 

Off course, it is not all about the specs, it is also about the look and feel of the mini-tablet and mainly about the ecosystem that you want to buy into: Google versus Amazon versus Apple.
I feel that these ecosystems are getting increasingly incompatible with one another and changing from one ecosystem to another can be painful and requires a lot of adaptation.

	Nexus 7	Kindel Fire	iPad mini
Screen Ppi (pixel per inch)	7" 1280x800 HD display at 216 ppi	7" multi-touch,  1024x600 resolution at 169 ppi or 1280x800 resolution for the HD option	7.9" multi-touch, 1024x768 resolution at 163 ppi
Front facing camera	1.2MP front-facing camera		1.2MP FaceTime HD camera
Weight in grams	340g	400g	308g 312 (wifi + cellular)
Memory	1 GB RAM	512MB	512 MB (???)
Battery	4325 mAh (Up to 8 hours of active use)	Up to 9 hours of reading, surfing the web on Wi-Fi, watching video, or listening to music.	Up to 10 hours of surfing the web on Wi-Fi, watching video or listening to music
CPU	Quad-core Tegra 3 processor	1GHz dual-core processor	Dual-core A5 but see here for full benchmarking
Size in mm	198.5 x 120 x 10.45	189 x 120 x 11.5	200 x 134.7 x 7.2
Wireless	WiFi 802.11 b/g/n + Bluetooth	Wi-Fi (802.11a/b/g/n); does not support connecting to ad-hoc (or peer-to-peer) Wi-Fi networks	Wi-Fi (802.11a/b/g/n; 802.11n on 2.4GHz and 5GHz) Bluetooth 4.0 technology
Connection	Micro USB	USB 2.0 (micro-B connector) port for connection to a PC or Macintosh computer or to connect to the Kindle PowerFast charging accessory	Ligthning conector
OS	Android 4.1 (Jelly Bean)	Android 2.3 Gingerbread OS	iOS6
Features	Microphone NFC (Android Beam) Accelerometer GPS Magnetometer Gyroscope	3.5 mm stereo jack and integrated stereo speakers	720p HD video FaceTime video calling over Wi-Fi or a mobile network3 Face detection Backside illumination iSight camera  5MP photos Autofocus Face detection Backside illumination Five-element lens Hybrid IR filter ƒ/2.4 aperture Siri - Use your voice to send messages, set reminders and more.
Price	£199 for 16GB and £159 for 6Gb	8 GB internal memory, approximately 5.5GB available for user content £129 or £169 for Fire HD with 16 GB or £209 32 GB on device	Wi-Fi 16GB £269 32GB £349 64GB £429 Wi-Fi + Cellular 16GB £369 32GB £449 64GB £529

Four years to digitise the phylogenies published so far!

"Monophyletic tree of organisms". Ernst Haeckel: Generelle Morphologie der Organismen, etc. Berlin, 1866. (Photo credit: Wikipedia)

Phylogenetic knowledge is being squandered at a rate of approximately 20,000 phylogenies a year (assuming that all papers with phylogen* in the title or abstract have illustrations of phylogenies). Fortunately, this loss of knowledge (and wasted money) is being tackled from multiple angles. On the one hand there is the open access movement that is striving to make publically-funded science freely available and accessible to everyone. For phylogenies this will hopefully go hand in hand with a greater submission of phylogenies to databases like TreeBASE. On the other, there are efforts to digitize past phylogenies: TreeRipper, TreeSnatcher and now TreeSnatcher Plus.

TreeSnatcher Plus has recently made a number of improvements on its predecessor and it was great to see that they benchmarked it against the same dataset as TreeRipper which contained phylogenies from the open access BMC Evolutionary Biology. They state that the average time for processing was 160s per phylogeny. I was interested to see how long it would take to digitise all the phylogenies produced to date. Assuming that all papers with phylogen* in the title has an image of a phylogeny, there are 734,585 published phylogenies according to ISI Web of Knowledge, that would require 4 years to digitise. The result might not be so bleak if Pubmed represents a more accurate number of phylogenies: 131,659 articles which would require a little under 1 year to digitise semi-automatically.

O.K. these numbers a pie in the sky but can we afford this wasted time?

Crowdsourcing science project for phylogenies?

Image via WikipediaThe idea behind crowdsourcing is that the answer to a question is often more likely to be correct if you average the answers from a large number of non-experts rather than a single expert in the field. The term "crowdsourcing" has also been used for projects that outsource repetitive or challenging work to a crowd via the internet.
I have been thinking of outsourcing the problem of conversion of embedded phylogenies in PDFs back to newick/nexus format and have been looking at various science projects that have used crowdsourcing.

The most impressive from my point of view is Galaxy Zoo which has already resulted in a number of publications and impressive discoveries. Astrophysicist use the crowd to categorise 1000s of galaxies and have expanded the crowd tasks to include matching images of galaxies with randomly simulated images.

Stardust@Home is another astrophysics project which asks that the crowd looks through images for dust particles brought back to earth by a spacecraft in 2006.

Another cool project is the Open Dinosaur Project which asks that the crowd aggregates published measurements of dinosaur limb bones for many different taxa from the literature and directly measured from specimens to study the evolutionary transitions from bipedality to quadrupedality.

Foldit is a computer game enabling the crowd to contribute to our understanding of how protein folds. Figuring out which of the many, many possible structures is the best one is regarded as one of the hardest problems in biology today and current methods take a lot of money and time, even for computers. The idea of using human's spare time to get further insight is genius!

Another game that might not be directly relevant to science is Google Image Labeler which I found rather addictive. Google gets users to label/tag images as a side-effect of playing a game and this is probably used to improve image searches on the web. I list it hear because I came across a few images of animals that in some cases were labeled down to the latin binomial.

UPDATE: An interesting new crowd sourcing project at http://www.oldweather.org/ to help gather information about past climates from hand written nautical records.

Share your trees and reduce your carbon footprint

I recently attended the SPDG in Glasgow. This is an discussion group on phylogenetics which takes place in various Universities across Scotland. The guest lecturer at the last meeting was Alexandros Stamatakis from Heidelberg. The main part of his talk was about the PaPaRa software which can be used to align short reads to a phylogeny. This is really useful for the identification of next-gen reads from environmental samples. However, he also talked about reducing the carbon footprint of computational biology by writing better algorithms and code. The effect of heavy computation on the environment was not new to me as I once sat through a video conference by Herve Philippe at the Entomological Society of America which was meant to be about "phylogenomics and the sister groups to Hexapoda" but ended up being about why he hadn't travelled to Reno, Nevada. He made valid points, which can be found here. At the SPDG, we also had a video conference from Erick Matsen and fortunately this time it was on topic. Erick is the organiser of phyloseminar which is well worth having a look at and could definitely lower your carbon footprint.

More efficient algorithms and programming, videoconferencing! This all got me thinking about the three Rs: REDUCE, REUSE and RECYCLE in the context of phylogenetics. We can all do our bit to REDUCE our carbon footprint when doing phylogenetics. For starters, is the analysis I want to do really necessary, does it have to run as long, can we use a better, more efficient algorithms. Secondly, we can REUSE the trees that others have already done but this means that we need to get much better at sharing our trees. TreeBASE and DataDryad are undoubtedly playing an important role in enabling us to share phylogenies and thus reduce our carbon footprint. However, as discussed in "Towards a taxonomically intelligent phylogenetic database" by Rod Page the pace at which we are publishing phylogenies is not being matched by the submissions to TreeBASE. This leaves us with the last option to RECYCLE our trees. This should only be a last resort but ends up happening most of the time. For this we need to get back to our raw materials, the sequences, which fortunately are more consistently shared in GenBank and redo the analyses.

Hopefully, this time round the algorithm will produce less carbon and the data will be submitted to TreeBASE!

Phylogeny digitisation

I was hunting around for further research on phylogeny image digitisation to see whether any advances had been made since I last published on the topic and to keep my previous post up-to-date. The main reason behind all of this is to see whether there would be a faster way to digitise a bunch of images that are accumulating on my hard drive. I thought it would be cool to do something with the ripped phylogenies for the iEvoBio Challenge but my current set of trees only has a total of 2,000 leaves and I need 10,000.
Anyway, I came across PHYLODIGM in my searches which looks promising. Thomas Laubach has also done some further work on TreeSnatcher Plus including using the benchmarking dataset from TreeRipper and a number of tree files found via Google searches. Additionally, he has released the source code under the GNU General Public License.
This all looks promising!!!

Not too far away from Gattaca

I once said jokingly to my brother that by 2020 we would be pretty close to being able to sequence a whole human genome the way they do it in the 1997 film Gattaca. Today with the press release from Oxford Nanopore it looks like we are one step closer. The most striking piece of kit is the cute MinIon, a USB stick you plug into your laptop and off you go sequencing. The sequencing approach and the concept of sequencing until you have found what you are looking for, is completely different to previous sequencing technologies. Forget about next-generation sequencing, this is revolutionary sequencing.
There's already a lot of coverage about Oxford nanopores press release and i found Nick Loman's blog posts on the new technology interesting.


"Run Until": DNA sequencing informatics on the GridION and MinION systems from Oxford Nanopore on Vimeo.