Playful coding: computing activities for schools

In many schools, computing is a topic that needs more encouragement. The Playful Coding project wants to make practical activities that can be run in schools to explore ideas in computer science. I've just been along to one of their meetings and seen it in action. It was extremely inspiring to be in a room full of people who didn't see running computing engagement activities as a chore, but as fun. They had all put a lot of thought into making fun activities and all wanted to run their activities with the groups of children.
 

It's an EU project involving teachers and university researchers from Spain, Romania, Italy, France and us in Aberystwyth, Wales. Each project partner had developed several activities and the purpose of the meeting was to tested out many of these activities on children and their teachers, and to start to develop a guide for teachers to explain how to use them. Until that guide is produced, you can still browse the activities and have a go with them. Try out for example:

• Collaborative story telling with Scratch, where each team of children makes a 30 second story with a character who moves from the left of the screen to the right. When all laptops are lined up, a character can move through every screen.
• Artificial Intelligence, where children get to think about whether a computer can ever be intelligence, and how would we know if it was?
• Poetry animation with Scratch, where children make characters that act out a poem.

There are lots more activities to choose from. Some just take an hour, some take a day, and some span a term. Some use robots, some use no particular equipment. They can be embedded into other lessons (languages, maths, science, art), or just standalone. And they are adaptable for different age ranges.

To follow the project see the Playful Coding website, follow #playfulcoding on Twitter or find Playful Coding on Facebook.

Seamless gene deletion

2015 is the year that genome editing really became big news. A new technique, "CRISPR/CAS", was named as Science magazine's breakthrough of the year as voted by the public from a shortlist chosen by staff.

However, people have been manipulating DNA through many useful methods long before CRISPR/CAS made headlines. Gene deletion is an important tool when trying to understand the function of genes. Take out a gene and see what effect it causes. Genes can be disrupted (by removing a portion of the DNA or inserting some extra DNA) or can be interfered with, for example via their RNA production, or they can be entirely deleted. It's common practice when removing a gene to insert a marker, so that we can easily select for the cells where this procedure has been successful. For example, to insert an antibiotic resistance gene as a marker, so that we can now grow the cells on a plate with an antibiotic. Then only those that have lost our gene of interest and gained antibiotic resistance will now grow. The trouble with this is that many gene deletions have no visible effect by themselves. If we also want to delete a second gene and a third, then we need more markers, or we need to be able to remove and reuse the marker we inserted. We also don't want the process to leave any scars behind that could destabilise the genome. We've just published a paper to help solve this problem.

This process of 'swap a gene of interest for a marker gene' can be achieved in many organisms by homologous recombination. This is a process used by many cells to repair broken strands of DNA. If we provide a piece of DNA that has a good region of similarity to the region just downstream of our gene of interest, and also a good region of similarity to the region just upstream of the gene of interest, but instead of the gene of interest, has the marker gene between these regions, then the normal cellular processes of homologous recombination will exchange the two. Some organisms perform homologous recombination very readily (S. cerevisiae for example). Others may need a little more encouragement, such as creating a double stranded break.

Our new paper A tool for Multiple Targeted Genome Deletions that Is Precise, Scar-Free and Suitable for Automation with Wayne Aubrey as first author uses a 3-stage PCR process to synthesise a stretch of DNA (a 'cassette') that will do everything. It will have good regions of similarity to the regions upstream and downstream of the gene of interest. It will contain a marker gene. And (here's the good bit), it will contain a specially designed region ('R') before the marker gene that is identical to the region that occurs just after the gene of interest. In this way, after homologous recombination has done its thing and inserted the DNA cassette instead of the gene of interest, there will be two identical R regions, one before the marker gene, and one after the marker gene. Sometimes the DNA will loop round on itself, the two R regions will match up and homologous recombination will snip out the loop, including the marker gene.

We can encourage this to happen and select for the cells that have had this happen if our marker is also 'counter-selectable'. That is, we'd like a marker for which we can add something to the growth medium so that now only cells without the marker will now grow. That is, we'd like to use a marker or marker combination for which we can first select for its presence and then counter-select for its absence. When we have this we can select for cells that have had the marker replace the gene, and then counter-select for cells that have now lost the marker too. So we have a clean gene deletion.

Of course we're always standing on the shoulders of giants when we do science. Our method is an improvement on a method by Akada 2006, so that no extra bases are lost or gained and the method requires no gel purification steps. Just throw in your primers and products and away you go. It's not fussy about quantity. No purification steps means that it could be automated on lab robots. And it could be used to delete any genetic component, not just genes. Give it a try!

Data science and a scoping workshop for the Turing Institute

In November I went to a workshop to discuss the remit of the Alan Turing Institute, the UK national institute for data science, with regard to the theme of "Data Science Challenges in High Throughput Biology and Precision Medicine". This workshop was held in Edinburgh, in the Informatics Forum, and hosted by Guido Sanguinetti.

The Alan Turing Institute is a new national institute, funded partly by the government, and partly by five universities (Edinburgh, UCL, Oxford, Cambridge, Warwick). The amount of funding is relatively small compared with that of other institutes (e.g. the Crick) and seems to be enough to fund a new building next door to the Crick in London, together with a cohort of research fellows and PhD students to be based in the new building. What should be the scope of the research areas that it addresses and how should it work as an institute? There are currently various scoping workshops taking place to discuss these questions.

Data science is clearly important to society, whether it's used in the analysis of genomes, intelligence gathering for the security services, data analytics for a supermarket chain, or financial predictions for the city. Statistics, machine learning, mathematical modelling, databases, compression, data ethics, data sharing and standards and novel algorithms are all part of data science. The ATI is already partnered with Lloyds, GCHQ and Intel. Anecdotal reports from the workshop attendees suggest that data science PhD students are snapped up by industry, ranging from Tesco to JP Morgan, and that some companies would like to recruit hundreds of data scientists if only they were available.

The feeling at the workshop seemed to be a concern that the ATI will aim to highlight the UK's research in the theory of machine learning and computational statistics, but risks missing out on the applications. The researchers who work on new and cutting edge machine learning and computational statistics don't tend to be the same people as the bioinformaticians. The people who go to NIPS don't go to ISMB/ECCB. And KDD/ICML/ECML/PKDD is another set of people again. These groups used to be closer, and used to overlap more, but now they rarely attend each others' conferences. Our workshop discussed the division between the theoreticians who create the new methods but prefer their data to be abstracted from the problem at hand, and the applied bioinformaticians, who have to deal with complex and noisy data, and often apply tried and tested data science instead of the latest theoretical ideas. To publish work in bioinformatics generally requires us to release code and data, and to have shown results on a real biological problem. To publish in theoretical machine learning or computational statistics, there is no particular requirement for an implementation of the idea, or to demonstrate its effectiveness on a real problem. There is also a contrast between the average size of research groups in the two areas. Larger groups are needed to produce the data (people in the lab to run the experiments, bioinformaticians to manage and analyse the data, and these groups are often part of larger consortia) whereas the theoreticians are often cottage-industry style research with just a PI and a PhD student. How should these styles of working come together?

Health informatics people worry about access to data: how to share it, get it, and ensure trust and privacy. Pharmaceuticals worry about dealing with data complexity, such as how to analyse phenotype from cell images in high throughput screening, having interpretable models rather than non-linear neural networks, and how to keep up with all the new sources of information, such as function annotations via ENCODE. GSK now has a Chief Data Officer. Everyone is concerned about how to accumulate data from new bio-technologies (microarrays then RNA-seq, fluorescence then imaging, new techniques for measuring biomarkers of a population under longitudinal study). Trying to keep up with the changes can lead to bad experiment design, and bad choices for data management.

There was much discussion about needing to make more biomedical data open-access (with consent), including genomic, phenotypic and medical data. There seemed to be some puzzlement about why people are happy to entrust banks with their financial data, and supermarkets with purchase data, but not researchers with biomedical data. (I don't share their puzzlement: your genetic data is not your choice, it's what you're born with, and it belongs to your family as much as it belongs to you, so the implications of sharing it are much wider).

All these issues surrounding the advancement of Data Science are far more complex and varied than the creation of novel and better algorithms. How much will the ATI be able to tackle in the next five years? It's certainly a challenge.

An executable language for change in biological sequences

A discussion on Twitter about whether there was a language for representing sequence edits prompted me to post my draft proposal for such a language. http://figshare.com/articles/Draft_proposal/1559009

Comments, criticism, collaboration and competition welcome. Hopefully I'll submit it shortly.

Notes from workshop on Computational Statistics and Machine Learning

I've just attended "Autonomous Citizens: Algorithms for Tomorrow's Society", a workshop as part of the Network on Computational Statistics and Machine Learning (NCSML). That's an ambitious title for a workshop! Autonomous Citizens are not going to hit the streets any time soon. The futuristic goals of Artificial Intelligence are still some way off. Robots are still clumsy, expensive and inflexible. But AI has changed dramatically since I was a student. Back in the days when computational power was more limited, AI was mostly about hand-coding knowledge into expert systems, grammars, state machines and rule bases. Now almost any form of intelligent behaviour from Google translation to Facebook face recognition makes heavy use of computational statistics to infer knowledge.

Posters: there were some really good posters and poster presenters who did a great job of explaining their work to me. In particular I'd like to read more about:

• A Probabilistic Context-Sensitive Model of Subsequences (Jaroslav Fowkes, Charles Sutton): a method for finding frequent interesting subsequences. Other methods based on association mining give lots of frequent but uninteresting subsequences. Instead, define a generative model, then go on to use data and EM to infer the parameters of the model. 
• Canonical Correlation Forests (Tom Rainforth, Frank Wood): a replacement for random forests that projects (a bootstrap sample of) the data into a different coordinate space using Canonical Correlation Analysis before making the decision nodes.
• Algorithmic Design for Big Data (Murray Pollock et al): Retrospective Monte Carlo. Monte Carlo algorithms with reordered steps. There are stochastic steps and deterministic steps. The order can have a huge effect on efficiency. His analogy went as follows: imagine you've set a quiz with a right answer and a wrong answer. People submit responses and you need to choose a winner. You could first sort them all into two piles (correct, wrong) and then pick a winner from the correct pile (deterministic first, then stochastic). Or you could just randomly sample from all results until you get a winner (stochastic first). The second will be quicker.
• MAP for Dirichlet Process Mixtures (Alexis Boukouvalas et al): a method for creating a Dirichlet Process Mixture model. This is useful as a k-means replacement where you don't know in advance what k should be, and where your clusters are not necessarily spherical.

Talks: these were mostly full talks (approx one hour), but then we had a short introduction to the Alan Turing Institute with Q&A at the end.

The first talk presented the idea of an Automated Statistician (Zoubin Ghahramani). Throw your time series data at the automated statistician and it'll give you back a report in natural language (English) explaining the trends and extending a prediction for the future. The idea is really nice. He has defined a language for representing a family of statistical models, a search procedure to find the best combination of models to fit your data, an evaluation method so that it knows when to stop searching, and a procedure to interpret/translate the models and explain the results. His language of models is based on Gaussian processes with a variety of interesting kernels, together with addition and multiplication as operators on models, and also allowing change points, so we can shift from one model combination to another at a given timepoint.

The next two talks were about robots, which are perhaps the ultimate autonomous citizens. Marc Deisenroth spoke about using reinforcement learning and Bayesian optimisation as two methods for speeding up learning in robots (presented with fun videos showing learning of pendulum swinging, valve control and walking motion). He works on minimising the expected cost of the policy function in reinforcement learning. His themes of using Gaussian processes, using knowledge of uncertainty to help determine which new points to sample were also reflected in the next talk by Jeremy Wyatt about robots that reason with uncertain and incomplete information. He uses epistemic predicates (know, assumption), and has probabilities associated with his robot's rule base so that it can represent uncertainty. If incoming data from sensors may be faulty, then that probability should be part of the decision making process.

Next was Steve Roberts, who described working with crowd sourced data (from sites such as zooniverse), real citizens rather then automated ones. He deals with unreliable worker responses and large datasets. People vary in their reliability, and he needs to increase accuracy of results and also use their time effectively. The data to be labelled has a prior probability distribution. Each person also has a confusion matrix, describing how they label objects. These confusion matrices can be inspected, and in fact form clusters representing characteristics of the people (optimist, pessimist, sensible, etc). There are many potential uses for understanding how people label the data. Along the way, he mentioned that Gibbs sampling is a good method but is too slow for his large data, so he uses Variational Bayes, because the approximations work for this scenario.

Finally, we heard from Howard Covington, who introduced the new Alan Turing Institute which aims to be the UK's national institute for Data Science. This is brand new, and currently only has 4 employees. There will eventually be a new building for this institute, in London, opposite the Crick Institute. It's good to see that computer science, maths and stats now have an discipline-specific institute and will have more visibility from this. However, it's an institute belonging to 5 universities: Oxford, Cambridge, UCL, Edinburgh and Warwick, each of which has contributed £5million. How the rest of us get to join in with the national institute is not yet clear (Howard Covington was vague: "later"). For now, we can join the scoping workshops that discuss the areas of research that are relevant to the institute. The website, which has only been up for 4 weeks so far, has a list of these, but no joining information. Presumably, email the coordinator of a workshop if you're interested. The Institute aims to have 200 staff in London (from Profs to PhDs, including administrators). They're looking for research fellows now (Autumn 2015), and PhDs soon. Faculty from the 5 unis will be seconded there for periods of time, paid for by the institute. There will be a formal launch party in November.

Next year, the NCSML workshop will be in Edinburgh.

ISMB/ECCB 2015

ISMB/ECCB 2015 (and HitSEQ 2015) was held in Dublin, just across the Irish Sea from us here in Aberystwyth. So off we went, to find out the latest research in bioinformatics. There were many parallel tracks, but the recurring themes of the talks I attended were:

• lots of work on human genomics, particularly disease, particularly cancer
• single cell analysis, finding variation (SNVs) from clonal populations, haplotype resolution
• sequencing technologies: RNA-seq, sequencing of methylation, Hi-C sequencing, ultra deep sequencing and lots of promise for long reads
• reference sequences: most people were working with a reference rather than de-novo
• training bioinformaticians, maintaining software, keeping a core of bioinfomaticians
• the Burrows Wheeler transform - does it solve every large-data problem?
• graphs, and ways of cleaning up graphs, adding weights to graphs, finding minimal/maximal components of graphs

There wasn't very much about the following topics, though they did make occasional appearances:

• text mining
• metagenomics
• multi-omics

Too hard? Dropping out of favour? Or perhaps people working in these areas just don't attend this particular conference?

Aberystwyth PhD students with their posters: Stefani Dritsa, Sam Nicholls, Tom Hitch, Francesco Rubino

The keynote talks tended to be of the kind that long-established group PIs do well. They're the "Here's a summary of all the work my group has been doing for the past 5-10 years to answer this particular biological question" talk. While I admire their determination and group size, I feel that they're speaking only to a subgroup of the audience with this kind of talk, and that a keynote should somehow also aim to more generally inspire the audience to go out and do great work, have new ideas, think in new directions, and not just to have learned a little more about that specific subject area. The far more off-the-wall non-keynote talk by David Searls about a bioinformatic analysis of James Joyce's book Ulysses fascinated the audience, and provided exactly that. He received a huge round of applause.


The jobs notice boards were full (below are just 2 of the notice boards). More bioinformaticians are clearly needed!

Many of the conference talks are now online, but you need to be a member of ISCB to see them http://www.iscb.org/ismb-mm/media-ismbeccb2015. The papers are also collected in a special ISMB/ECCB issue of Bioinformatics.

Aber Bioinformatics Workshop

Last week we had the 2nd Aber Bioinformatics Workshop. It's an internal workshop for work-in-progress talks, posters and networking and the aim is for us all to keep up with what's going on in Aberystwyth in bioinformatics across departments and institutes. We had a wide range of talks on genomics and sequence analysis, metabolomics, optimising proteins, population and community modelling, data infrastructure and other topics. Here's the programme for the day.

Photo of all the attendees, taken by Sandy Spence

It was great to see that we now have so many people interested and working in bioinformatics, despite the difficulties in trying to understand all sides of the story (the biology, the computing, the statistics, etc). We talked about the range of modules and courses that were available to help people get up to speed with this, and how we should do more to let new PhD students know what is available. Also, now that we've had the workshop, hopefully we're more aware of the expertise and facilities available here in Aber, so we now know who to approach with questions and ideas.

At the end of the day we moved down to the pub, and continued to discuss more random topics: beetles, plant senescence, hens, temperature sensing wires for computer clusters, and concordance in Shakespeare texts. I'm sure this all helps in the long run.