It's no secret that we're huge fans of Python at our consulting firm Amber Biology. At least 90% of all our consulting projects involve some Python coding and it's such a versatile, productive and expressive language that we like to call it "The Swiss Army Knife of programming languages". One of the things we commonly hear about Python however, goes something like this:

"They say that Python is too slow and inefficient to use for heavy computational tasks because it's an interpreted language."

Is this the same "they" that also say you can develop extraordinary night vision by eating lots of carrots?

At Amber Biology we have used Python for many computationally intensive research problems, for example, simulating the use of a novel next-generation sequencing laboratory protocol on the human genome.

Yes, the standard Python distribution may not be as fast under certain circumstances as some other languages that are compiled (like C for example), but there are many approaches to using Python in ways that can raise its performance to a level that is at least in the same ballpark as these compiled languages, if not equal to them.

In addition to being written in the Python language itself, Python modules can also be written in other languages such as C and then compiled to machine executable instructions and linked to the Python interpreter. This means that your Python code can be running as interpreted code, or directly as machine-executable instructions - all from within the same application.

The official Python interpreter is actually written in C, as are many of the modules that come with it as part of the Python Standard Library. This means that when you call these standard library modules from your code that is running in the Python interpreter, the functions and methods that these modules contain can be run directly (and very efficiently) as compiled machine-executable instructions for as long as they are able to run without encountering a reference to anything back in the interpreter loop. Essentially what this means is that if you structure your code well, you can get something akin to compiled language performance, even from an "interpreted" language like Python.

Don't believe us?

Researchers working in the fields of next-generation sequencing and genomics, routinely work with sequences that number in the millions or even billions of bases. As an example of how rapidly and efficiently Python can work when applied to these kinds of problems, let's take the largest contiguous piece of the human genome, Chromosome 1, and use Python to search its approximately 250 million bases for potential gene promoter sequences.

Gene promoters often consist of relatively conserved recognition sequences separated by a spacer sequence, often of variable length. Many bacterial gene promoters for example, consist of two distinct recognition sequences upstream of the gene to be transcribed like this:

5' ----PPPPPP-------------------PPPPPP----GGGGGGGGGGG … 3'
       |                        |         |
      -35                     -10         start of gene

The consensus sequence for the recognition site at the -10 position is tataat while the -35 site has a consensus sequence of ttgaca. We can quite easily craft a Python regular expression that would search a genome for such consensus promoter regions like this:

promoter = 'ttgaca...................tataat'

This would match any sequences that consisted of exactly the two consensus recognition sites separated by a fixed number of bases (with any of the 4 bases being allowed at those positions).

But what if the spacer between the consensus sequences is of variable length?

It turns out that regular expressions give us a simple and powerful way to deal with this situation as well. We can create search patterns that allow a variable number of bases between the recognition sites like this:

promoter = 'ttgaca.{15,25}tataat'

The curly brackets after the period are used to indicate a repeat of between 15 and 25 occurrences of the character preceding the brackets - in this case the period that matches any base. This search pattern would match any stretch of the genome consisting of the two consensus recognition sites, separated by between 15 and 25 bases of any kind.

So now we have a gene promoter pattern to search for, let's use Python to generate a synthetic Chromosome 1 - especially since this is just a computational performance test and it saves us all the time of having to find a Chromosome 1 file from an online human genome repository, and write the code to read and parse it. To create our synthetic chromosome, we'll simply use one of the methods in the Python random module to generate a randomized sequence of 250 million bases, like this:

import random
bases = ['a', 't', 'c', 'g']
sequenceList = []
for n in range(0, 250000000):
    sequenceList.append(random.choice(bases))
chromosome = ‘’.join(sequenceList)

This actually takes a few minutes on my iMac - much (much) longer than the time it will take to search this synthetic chromosome for sequences that match our consensus gene promoter pattern, as we shall see in a minute ...

So for this performance test we will need to use the Python time module, which as the name implies, handles time. The time.time() method returns the number of seconds since an arbitrary fixed date and time referred to in the Python documentation as the start of the epoch. This all sounds very prehistoric and/or apocalyptic, but you might be surprised to learn that this grand dawn of history is actually January 1st 1970.

Really?

This has to do with the fact the standard approach to counting time on computers was defined by the engineers of Unix which was one of the first real computer operating systems, and it has been inherited by virtually all other computer platforms as a kind of de facto standard for time.

And - since we are going to use Python regular expressions, we will also need to import the Python re module, from which we will use the wonderfully versatile re.finditer() method to search for our consensus gene promoter sequences. Instead of just returning matching sequences, the finditer() method returns an iterable collection of Python MatchObject objects, that allows the search results to be analyzed in more detail (for example - the start and end positions of the matching sequences are also recorded).

When we put it all together, the code to search our synthetic Chromosome 1 for consensus gene promoter sequences looks like this:

import time, re
promoter = 'ttgaca.{15, 25}tataat' 
t1 = time.time() 
result = re.finditer(promoter, chromosome) 
t2 = time.time() 
print 'Search time was', (t2-t1), 'seconds'

So let's run this Python code and see what we get ...

Search time was 0.00100994110107 seconds

Wow!

On my modest iMac desktop (circa 2014), it takes Python only about one millisecond to search the 250 million base sequence of Chromosome 1 for this consensus gene promoter pattern!

By extrapolation therefore, we could expect that excluding the time it would take to load and read the individual chromosome sequence files of the human genome - it would take only about 15 milliseconds to search its entire 3 billion bases for this gene promoter pattern!

By the way - just in case you were wondering if it was fast only because there was nothing for it to find, let's add a little additional code to print out the details of the matches found. We will use the data stored in the Python MatchObject to print the start and endpoints of each matching sequence, as well as the matching sequence itself (and to save space, we will only show the first 3 and last 3 matches here - but you get the idea). Finally, we will also print the total number of matching gene promoter sequences that were found in our synthetic chromosome.

nmatches = 0 
for match in result:
     nmatches += 1     
     print match.start(), match.end(), match.group() 
print 'Number of search hits = ', nmatches

And when we include this additional code, we get this ...

1199566 1199603 ttgacactcacatcatcagagccccacatagtataat
2103278 2103308 ttgacacacacagggtttgtgatttataat
3702112 3702141 ttgacactctttcaaaccaggactataat
 … 
 …
245627316 245627350 ttgacaaggtctccgtggccccggctattataat
246256184 246256220 ttgacaggattcctctcgttaattacatcgtataat
248653641 248653674 ttgacaaccgggctcgtaacgtattagtataat
Number of search hits =  185

So there you have it.

Next time somebody tells you that Python is too slow for heavy computation because it's an interpreted language - just politely tell them that you heard of a bridge in New York state that's up for sale at a bargain price!

© 2018 The Digital Biologist

In a previous article we posted on “The Digital Biologist”, we gave a very brief and simple introduction to Bayes’ Theorem, using cancer biomarkers as an example of one of the many ways in which the theorem can be applied to the evaluation of data and evidence for research. Bayesian analysis can be a really useful and practical computational tool in life science R&D, and in this brief article we will develop and use a really tiny snippet of Python code to demonstrate its application to biomarkers.

Let’s suppose that we are a company determined to develop a more reliable biomarker than CA-125 for the early detection of ovarian cancer. One direction we might pursue is to identify a biomarker that predicts disease in actual sufferers with a higher frequency i.e. a biomarker with a better true positive hit rate. We saw in the introductory article, that CA-125 only predicts the disease in about 50% of sufferers for stage I ovarian cancer and about 80% of sufferers for stage II and beyond. One of the dilemmas faced by physicians working in the oncology field, is that biomarkers like CA-125 can be poorly predictive of the disease in the early stages when the prognosis and options for treatment are better. It’s disheartening for both the patient and the physician to be able to get a reliable diagnosis only when the disease has already progressed to the point at which there are fewer good options for treatment.

Ovarian cancer affects about 1.5% of the population and the CA-125 test is predictive of the disease in about 80% of sufferers, which sounds pretty good right? Most people would probably look at those numbers and conclude that a positive test for the biomarker was associated with something like an 80% chance of having the disease.

Not so fast!

The CA-125 test also produces false positives in about 4% of patients without the disease  (i.e. the test result indicates disease where none is actually present). Armed with these numbers, we might feel that CA-125 has a pretty bright future in the clinic based upon the following reasoning – it will only fail to detect the disease in about 2 out of 10 sufferers and it will only produce a misdiagnosis of the disease in about 4 out of 100 healthy patients.

Once again – not so fast.

Let's develop a very simple Bayesian model with just a few lines of Python, and then we can plug in these numbers and see what we get.

We'll write a very simple Python function that takes as input, the rate of occurrence of the disease in the population, the rate of true positive test results, and the rate of false positive test results. The function returns the actual probability of having the disease given a positive test result. Using the discrete Bayesian formula shown here, we will be formulating the problem like this:

• p(B) = probability of having the disease
• p(A) = probability of a positive test result
• p(B|A) = probability of having the disease given a positive test result
• p(A|B) = probability of getting a positive test result given that you have the disease

Here's the Python function:

def biomarker(p_disease, p_true_positive, p_false_positive):     p_healthy = 1.0 - p_disease     p_positive = p_true_positive * p_disease + p_false_positive * p_healthy     return (p_true_positive * p_disease) / p_positive

Firstly you will notice that in order to calculate the denominator p(A) in the discrete Bayesian equation which corresponds to the probability of a positive test result, we need to calculate the cumulative probability of getting any positive test result - in other words, the total probability of a positive test result for healthy and diseased patients. The probability of being healthy (i.e. not having the disease), is simply 1.0 - p(B). Once we have this value (p_healthy in the Python function), we can then calculate the cumulative probability of any kind of positive test result. We already know p(A|B) the probability of getting a positive test result when you have the disease, because we passed this to the function as p_true_positive. Similarly, we also know p(B), the probability of having the disease since this is also passed to the function as p_disease. Once we have compiled the cumulative probability of getting any kind of positive test result (p_positive), we are ready to deliver the result.

You can actually test  the numbers for yourself in this live Python code window below. At the prompts, just enter the fractional occurrence rate of your disease in the population, and the fractional rates of true and false positives, to calculate the probability that a patient has the disease given a positive result from your test.

<span data-mce-type="bookmark" style="display: inline-block; width: 0px; overflow: hidden; line-height: 0;" class="mce_SELRES_start"></span>

Let's plug in the real numbers for the CA-125 test for ovarian cancer ...

• p_disease (p(B)) = 0.015
• p_true_positive = 0.8
• p_false_positive = 0.04

and when we run the code we get ...

Probability of disease given a positive test result = 0.23

Wow!

There is only a probability of about 23% of actually having ovarian cancer given a positive CA-125 test - in other words, the patient who had the positive test result has only about 1 chance in 4 of actually having the disease, based upon the positive test result.

Let's come back to our hypothetical diagnostic company seeking to develop a better ovarian cancer test and let's say that the company's scientists decide that they have two potential approaches - increase the true positive rate of the test - or decrease the false positive rate.

So imagine that our hypothetical diagnostic company has made a heavy R&D investment in identifying a biomarker with a much better rate of true positives. If our new biomarker has true positive rate of 95% (a fairly significant improvement on our previous value of about 80%) and the same roughly 4% false positive test rate as previously, how much better off are we?

If we plug the numbers into our Bayesian model, the answer is “not much”

Probability of disease given a positive test result = 0.27

Even with such a dramatic improvement in the rate of true positives, the new prediction is barely any more accurate than the old one.

What if instead of pursuing a better true positive hit rate, our company had invested in reducing the false positive test rate?

Without altering the true positive rate of about 80%, let's imagine that they are able to reduce the biomarker’s false positive rate from about 4% to 1%.

Probability of disease given a positive test result = 0.55

Much better but still not great. What about a false positive rate of 0.5%?

Probability of disease given a positive test result = 0.71

The Bayesian model clearly tells us in the case of ovarian cancer - that our hypothetical company is much better off investing its R&D dollars in the pursuit of lower false positive test rates rather than higher true positive test rates. Even a 99% true positive test rate barely shifts the probabilities associated with a positive test result, whereas getting the false positive test rate down to 1% improves the probability of a true diagnosis from less than 1 in 4, to better than even. Even this scenario however, is far from ideal.

If you look at the actual numbers in the model with regard to the populations of tested patients with and without the disease, there is another valuable lesson to be learned, and it is one that illuminates the reason why improving the true positive test rate while ignoring the false positive test rate is less than optimal.

It is the overwhelmingly larger population of healthy patients versus those with the disease, that is skewing the probability numbers against us - and the lower the incidence of the disease, the worse this problem will be.

If ovarian cancer had a higher incidence of say, 1 in 10 women instead of 1 in 72 as is actually the case, a positive test result with CA-125 would correspond to an almost 70% probability of the patient actually having the disease. By contrast, if the ovarian cancer incidence was 1 in 1000 women, a positive test result with CA-125 would still correspond to less than 1 chance in 50 of the patient actually having the disease.

The lower the incidence of the disease you want to diagnose, the correspondingly lower your false positive test rate needs to be.

© 2018, The Digital Biologist

A lot of very happy people on the NASA - National Aeronautics and Space AdministrationTESS mission team today. At 6:51PM ET the TESS spacecraft was successfully launched from Cape Canaveral on a SpaceX Falcon 9 booster and carried into orbit. As of the time of writing, it has just successfully deployed its solar arrays and is broadcasting telemetry to mission control. Wow!

My partner and I at Amber Biology are very happy and proud to have been given the opportunity to play a very tiny part in this amazing venture as contractors on the mission.

With apologies for the quality of the photos which were snapped from a TV during the live stream.

© 2018, The Digital Biologist

Our consulting firm Amber Biology was recently commissioned by a biotechnology company, to build a dynamic, agent-based model of the biological pathways that the company's scientists were studying as part of their research and development program. As I was presenting our model to the company's researchers, I could sense their anticipation at seeing, perhaps for the first time, the concerted activities of all of the genes, proteins and other molecules that they had been studying in their research. It's something of an epiphany to witness what was essentially a bunch of data points, suddenly organized and integrated into a biological system whose behavior is an emergent property of all that data.

With the model in hand, we were immediately able to do some very interesting, virtual "what-if" experiments with it. What if the activity of this regulatory protein was only at 50% of its level in normal, healthy cells?How would the response of this pathway be different in a cell where this gene was deficient or knocked out? Moreover, as the discussion developed around the results from the model, some of the less commonly considered functions of models (beyond their ability just to make predictions) started to come to the fore - the generation and testing of hypotheses, the conception and development of new laboratory experiments and the ability of the model itself to serve as a vehicle for communication and the discussion of ideas. Significant insights can be gained from having a working, dynamic model of the biological pathways that you are studying. Within the space of the (approximately) one hour that my presentation to this company lasted, we were able to demonstrate (among other things) the detailed chain of cause-and-effect by which a clinically-observed deficiency in a particular regulatory molecule can lead to a disease state; and also that the model correctly predicts a biomarker for this disease state derived from the relative levels of production of certain metabolites generated by the pathway.

From the perspective of providing answers to biological questions, the data that you need to build such a model are exactly that - just data. Without some kind of conceptual model with which to reason about them, biochemical parameters such as molar concentrations, kinetic rates for the binding, unbinding and enzymatic modification of molecules etc. do not really constitute "knowledge" at the biological level any more than measurements of the position of Mars constitute "knowledge" at the astronomical level. Framed within the context of a model however, these observations can become a foundation for reasoning about the system at a higher, biological level - moving the conversation from the microscopic domain of molecules, kinetic rates, affinities and so on - to considerations of cellular responses, biological dysfunction and disease states.

In this era of automated, digitized laboratory instruments, and data storage on the cloud, there is certainly no shortage of biological data - yet the investment by most life science research organizations in the kind of infrastructure or expertise needed to turn data into real knowledge, has certainly not kept pace with their ability to generate new data. To be perfectly clear - the data are valuable, but data alone seldom yield any kind of actionable insights that would for example, assist anyone trying to develop a therapeutic intervention for a biological system. Rather than being organized and integrated into conceptual models that could explain it, the real value of the data is often buried under mountains of well-meaning but fruitless data analytics and data visualization. For the kind of complex biological systems that life scientists study - the kind of systems that produce big data sets whose complexity mirrors their own - it’s a long shot to hope to draw much in the way of meaningful biological insight from some purely mathematical fitting, filtering, and visualization of such data.

Computational modeling is certainly not in the mainstream of life science research in the way that it is in other fields such as physics and engineering, and a lot of this probably has to do with the difficulty of overcoming the challenges posed by the sheer complexity of biological systems. The most commonly used dynamic modeling paradigms for modeling biological systems, such as ordinary differential equations (ODEs) for example, have largely been borrowed from these other fields where they have enjoyed great success. Their application to biology however is more problematic - principally because of their inability to adequately represent the aforementioned complexity of the systems that biologists would like to model, but also because of the disconnect between the causal descriptions that biologists typically use to describe the systems they study, and the more abstract (and opaque) idioms used to construct and query such models. Despite this however, more recent approaches that include agent and rule-based modeling platforms developed specifically for biology are starting to mature to the point at which we may soon witness a much greater level of adoption of computational modeling in the mainstream of life science research, and a much wider recognition of its role as an invaluable adjunct to laboratory-based research and development programs.

© The Digital Biologist 2018

At Amber Biology we are very excited to start teaching our coursePython For Life Scientistsat Boston Children's Hospital on Feb. 13th. This 8-week course is a combination of classroom teaching and workshop sessions, and is based upon our book Python For The Life Sciences. Like the book, we created this course specifically for biologists with little or no prior experience writing code.

At most colleges unfortunately, courses in coding and basic computer science are still not a part of the core curriculum for the life sciences, yet biologists are increasingly discovering the need to move beyond calculators and spreadsheets in order to handle their data and the more quantitative aspects of their own research. Our aim on this course, is to provide the life scientist with a foundation in Python sufficient for them to be able to start using it in their own research.

Unfortunately this current session of the course has already received about twice the number of applicants than we have places available, so enrollment is closed. If however you would like to have this course taught on-site at your organization, please contact us so that we can schedule a class for you.

© The Digital Biologist 2018

A while back, I published an article Big Data Does Not Equal Big Knowledge, Fast forward a year or two and it's still the case that the tangible results that Big Data solutions have yielded in the life sciences, pale by comparison with the investment in them. So long as life science research organizations are happy to keep paying for Big Data solutions that generate more fancy infographics than actual insights, vendors of Big Data solutions are happy to keep selling them. It's going to be interesting to see how long this will continue as some of the big players like IBM are also heavily pushing their Big Data/AI solutions like Watson. To date, it seems like little has changed and the original article is as relevant today as it was when I published it.

© 2017, The Digital Biologist

One of the strengths of this platform therefore, was that in addition to being able to model the dynamic behavior of the assembled system, it also allowed for the kind of powerful static analysis that can provide analytical answers to difficult questions of this sort.

Everything you see in the brief demo video below, is running live in a web browser. I hope you enjoy it.

© 2017, The Digital Biologist

I was recently interviewed by Leanpub co-founder Len Epp for their author podcast, and chat with Len about my career, books, and about the publication on the Leanpub platform, of our book Python for the Life Sciences by my consulting firm Amber Biology.

"I was completely hooked from the minute I started writing BASIC on a home computer. All through college, I always pursued projects and electives where I had a chance to do computing. So that’s always been a big part of my career too. Biophysics is a very computational, quantitative, numerically-intensive field, and so the computer stuff has always played a very large part in that."

You can read and listen to the full interview at Leanpub's blog site.

© 2017, The Digital Biologist

Our consulting firm Amber Biology is very excited to have been hired as a contractor for the NASA TESS (Transiting Exoplanet Survey Satellite) mission. Due to be launched from Cape Canaveral as early as March 2018 on one of SpaceX's new Falcon 9 reusable rockets, the TESS spacecraft will search for exoplanets orbiting distant stars, that could potentially be earth-like worlds on which liquid water could exist, and which may also have the right kind of conditions to support life. Although most of the computational work we are doing is in the field of astrophysics, this is an origin of life project and therefore still arguably biology!

Amber Biology is involved in the creation of code involved in both the imaging processing for the spacecraft itself, and the data analysis pipeline that scientists on the ground will use to process the images that the spacecraft sends back to earth on each closest approach of its orbit.

© 2017, The Digital Biologist

Facebook now sports a very nice feature that allows you to make short videos out of simple slideshows. Here is the one I made for our new book Python For The  Life Sciences, that we published recently.

© The Digital Biologist

If you’re interested in the future of research and scholarship, and like many in the field, you also subscribe to the consensus view that the current system is broken, you won’t want to miss the Ronin Institute’s “The Future of Careers in Scholarship”, being held in Cambridge MA on November 5th. The unconference format of the meeting, will even allow you and other attendees to shape the agenda of the meeting, so come prepared to be an active participant rather than just a spectator. The meeting will be hosted at The Democracy Center in Harvard Square and you will have the chance to meet and network with an eclectic and forward-thinking group of people from a range of different research areas.

If you are really interested in the future of research and scholarship, and would like to get involved in the movement to advance beyond our current broken system and build new models for doing research, take a look at the Ronin Institute website. The Ronin Institute is devoted to facilitating and promoting scholarly research outside the confines of traditional academic research institutions.

© The Digital Biologist

The book does however cover an incredible range of life science research topics from biochemistry and gene sequencing, to molecular mechanics and agent-based models of complex systems. We hope that there's something in it for anybody who's a life scientist with little or no computer programming experience, but who would love to learn to code.

[embed]https://www.youtube.com/watch?v=V7eH0SJ3kfw[/embed]

For the latest news on the book, including all the free revisions and updates that are included with any purchase of the book, sign up for the (zero spam) Python For The Life Sciences Mailing List

© The Digital Biologist

There's hardly a life science lab you can walk into these days, without seeing a ton of 96-well plates and instruments that read and handle them. That's why we've dedicated an entire chapter of our forthcoming book Python For The Life Sciences, to the humble 96-well plate. The chapter introduces the use of Python for handling laboratory assay plates of many different sizes and configurations. It shows the reader how to read plate assay data from files formatted as comma-separated values (CSV), how to implement basic row and column computations, how to plot multi-well plates with the wells color-coded by their properties, and even how to implement the high level code necessary for driving instruments and robots through devices like Arduinos.

And this is just one of about 20 chapters designed to introduce the life scientist who wants to learn how to code, to the wonderful and versatile Python programming language.

Almost all of the code and examples in the book are biology-based and in addition to teaching the Python programming language, the book aims to inspire the life scientist reader to bring the power of computation to his or her research, by demonstrating the application of Python using real-world examples from across a wide range of biological research disciplines.

The book includes code and examples covering next-generation sequencing, molecular modeling, biomarkers, systems biology, chemical kinetics, population dynamics, evolution and much more.

Python For The Life Sciences should be available as an eBook this fall (2016), so if you're a life scientist interested in bringing a computational skill set to your research and your career, visit the book's web page and sign up to our (no spam) mailing list for updates about the book's progress and publication.

© The Digital Biologist

Here at The Digital Biologist, we have reported at some length on the Theranos scandal, since it serves as an example of the way that biological data can be used (and misused) in the development of new healthcare technologies. Now the KQED Science web site has published a wonderful cartoon history of the rise and fall of Theranos, along with links to the major articles that described this history as it was unfolding. It's a kind of an everything-you-ever-wanted-to-know-about-the-Theranos-scandal-but-were-afraid-to-ask, all presented on a single page. Enjoy :-)

© Digital Biologist

© The Digital Biologist

To be fair, it always was a 'thing', but now that The Future of Research has made it official by becoming a fully-fledged non-profit organization, it's now really a 'Thing' (with a big T). This is no small feat for a group that started out as a loosely-knit assembly of grassroots activists who share a common interest in improving things for academic researchers. In only three years, they have built both national  and international recognition for their movement. They were named 2015 People of the Year by Science Careers, as well as landing a 2-year, $300,000 grant from the Open Philanthropy Project, to help them in their work in assisting junior scientists in grassroots efforts to change science policy. So congratulations to the Future of Research on becoming a Thing with a big T. Anyone who cares about the future of academic research and the working conditions and job prospects of those who pursue careers in it, should consider joining this very important conversation that the Future of Research has started.

As history has taught us over and over, systems that are broken, dysfunctional and unfair, rarely if ever transform or dismantle themselves from the top down, but rather, from the bottom up. Those who benefit the most from such systems are also usually its gatekeepers, and they will generally strive (consciously or unconsciously) to preserve it since they potentially have the most to lose from changing it. This is precisely why grassroots organizations like The Future of Research are so important as instruments of social change.

© The Digital Biologist

... the life science Big Data scene is largely Big Hype. This is not because the the data itself is not valuable, but rather because its real value is almost invariably buried under mountains of well-meaning but fruitless data analytics and data visualization. The fancy data dashboards that big pharmaceutical companies spend big bucks on for handling their big data, are for the most part, little more than eye candy whose colorful renderings convey an illusion of progress without the reality of it.

Read the full article on LinkedIn.

© The Digital Biologist

Our book Python For The Life Sciences is now nearing publication - we anticipate sometime in the early summer of 2016 for the publication date. Quite a number of people have asked us to release a table of contents for the book, so without further ado, here is the first draft of the table of contents. If you would like to receive updates about the book, please sign up for our book mailing list.

Python at the bench: In which we introduce some Python fundamentals and show you how to ditch those calculators and spreadsheets and let Python relieve the drudgery of basic lab calculations (freeing up more valuable time to drink coffee and play Minecraft)

Building biological sequences: In which we introduce basic Python string and character handling and demonstrate Python’s innate awesomeness for handling nucleic acid and protein sequences.

Of biomarkers and Bayes: In which we discuss Bayes’ Theorem and implement it in Python, illustrating in the process why even your doctor might not always estimate your risk of cancer correctly.

Reading, parsing and handling biological sequence data files: Did we already mention how great Python is for handling biological sequence data? In this chapter we expand our discussion to sequence file formats like FASTA.

Regular expressions for genomics: In which we show how to search even the largest of biological sequences quickly and efficiently using Python Regular Expressions - and in the process, blow the lid off the myth that Python has to be slow because it is an interpreted language.

Biological sequences as Python objects: Just when you thought you had heard the last about sequences, we explore the foundational concept of Object Oriented Programming in Python, and demonstrate a more advanced and robust approach to handling biological sequences using Python objects.

Slicing and dicing genomic data: In which we demonstrate how easy it is to use Python to create a simple next-generation sequencing pipeline - and how it can be used to extract data from many kinds of genomic sources, up to and including whole genomes.

Managing plate assay data: In which we use Python to manage data from that trusty workhorse of biological assays, the 96-well plate.

Python for structural biology and molecular modeling: In which we demonstrate Python’s ability to implement three-dimensional mathematics and linear algebra for molecular mechanics. It’s nano but it’s still biology folks!

Modeling biochemical kinetics: In which we use Python to recreate what happens in the biochemist’s beaker (minus the nasty smells) - as well as using Python to model the cooperative binding effects of allosteric proteins.

Systems biology data mining: In which we demonstrate how to parse and interrogate network data using Python sets, and in the process, tame the complex network “hairball”.

Modeling cellular systems: In which we introduce the Gillespie algorithm to model biological noise and switches in cells, and use Python to implement it and visualize the results - along with some pretty pictures to delight the eye.

Modeling development with cellular automata: In which we use the power of cellular automata to grow some dandy leopard skin pants using Turing’s model of morphogenesis with Python 2D graphics. Note to our readers: no leopards were harmed in the writing of this chapter.

Modeling development with artificial life: In which we introduce Lindemeyer systems to grow virtual plants and use Python’s implementation of Turtle LOGO. Don’t worry, these plants will not invade your garden (but they might take over your computer).

Predator-prey dynamics in ecology: In which we let loose chickens and foxes into an ecosystem and let ‘em duke it out in a state-space that is visualized using Python’s animation features.

Modeling virus population dynamics with agent-based simulation: In which we create a virtual zombie apocalypse with agents that have internal state and behaviors. These are definitely smarter-than-usual zombies that illustrate an approach in which Python’s object-oriented programming approach really shines.

Modeling evolution: In which we use the Wright-Fisher model to demonstrate natural selection in action, and show how being the “fittest” doesn’t always mean that you will “win”. Think Homer Simpson winning a game of musical chairs.

© The Digital Biologist

There's a great deal more than money at stake here.

This week's regulatory call for Theranos could literally make or break the company depending upon which way it goes. Failure to be in regulatory compliance could bring with it, a host of new problems for the struggling company including crippling fines and the inability to operate until such time as it can demonstrate that it has addressed the problems raised by the regulatory agencies. Most damaging of all however, this would be yet another huge blow to its already strained credibility with investors and healthcare provider partners, some of whom have already withdrawn or suspended their relationships with Theranos over doubts about the efficacy and accuracy of its blood tests. Beyond the problems that this could create for Theranos itself, many industry observers fear the effects that it could have on the entire sector if a contagion of doubt and panic were to grip investors and financiers, potentially stemming the flow of biotechnology venture funding and capital.

© The Digital Biologist

In conjunction with my business partner Alex Lancaster, we are very excited for this early release of a sample chapter from our forthcoming book Python For The Life Sciences. This book is written primarily for life scientists with little or no experience writing computer code, who would like to develop enough programming knowledge to be able to create software and algorithms that they can use to advance or accelerate their own research. These are probably scientists who are currently using spreadsheets and calculators to handle their data, but who have probably promised themselves that at some point when the opportunity arises, they will learn to write code. If this pretty well describes your situation, then your wait is over and the opportunity is knocking. This could very well be just the book you have been waiting for!

In short, this is the book that would like to have read when we were learning computational biology.

The aim of this book

The aim of this book is to teach the working biologist enough Python that he or she can get started using this incredibly versatile programming language in their own research, whether in academia or in industry. It also aims to furnish a Python foundation upon which the biologist can build by extrapolating from the broad set of Python fundamentals that the book provides.

What this book is not

This book is not another comprehensive guide to the Python programming language, nor is it intended to be a Python language reference. There are already plenty of those out there, and easily accessible online. For this reason, you will find that there are many (many) aspects and areas of the Python language that are not covered. In a similar vein, this book is not intended to be a life science primer for programmers and computer scientists.

A tour of computational biology beyond bioinformatics

This book is all about using computational tools to jumpstart your biological imaginations. We will show the reader the range of quantitative biology questions that can be addressed using just one language from a range of life sciences. The examples are deliberately eclectic and cover bioinformatics, structural biology, systems biology to modeling cellular dynamics, ecology, evolution and artificial life.

Like a good tour, these biological examples were deliberately chosen to be simple enough not to impede the reader’s ability to assimilate the Python coding principles being presented - but at the same time each scientific problem illustrates a simple, yet powerful principle or idea. By covering a wide variety of examples from different parts of biology, we also hope that the reader can identify common features between different kinds of models and data and encounter unfamiliar, yet useful ideas and approaches. We provide pointers and references to other code, software, books and papers where they can explore each area in greater depth.

We believe that exploring biological data and biological systems should be fun! We want to take you from the nuts-and-bolts of writing Python code, to the cutting edge as quickly as possible, so that you can get up and running quickly on your own creative scientific projects.

The sample chapter shows how to use Python to mine and understand data from transcription factor networks and you can get it here.

© The Digital Biologist