Neuromancy Blog

I’m on the move!

Neuromancy — Fri, 14 Jan 2011 13:34:58 +0000

I’ve moved house! I’ve just seen that I’ve been officially launched over at the Southern Fried Science blogging network. So ladies and gentlemen, update your bookmarks and adjust your RSS readers. I can now be found at http://neuromancy.southernfriedscience.com

I’ve locked comments on all the articles here, but everything’s been transferred over to SFS, comments and all, so you can carry on any discussions over there.

Go forth and comment!

Craig

Anscombe’s quartet and the importance of graphing your data

Neuromancy — Thu, 30 Dec 2010 23:52:11 +0000

I’ve stalled a little bit in my thesis writing – I’ve only got reanalysis and in-depth rewriting to be done on this chapter, which I’m not keen to do at this time of night. So I thought I’d bash out a quick post before the new year.

When dealing with sets of data from familiar experiments, it might be tempting to throw the numbers into your favourite statistics software package, and report the coefficient and p value. But researcher beware! Strange things may be hiding in your data… Anscombe’s quartet is a fine example of this. The quartet is four sets of data that have the same sample statistics (mean, variance, correlation coefficient and regression equation), but when graphed, they are clearly very different.

Anscombe's quartet plotted, from Wikipedia

The quartet is only an illustrative example of what is possible; the Wikipedia article has links to other similar data.

But graphing your data doesn’t just guard against mistakes, it can also allow you to see patterns in your data that you hadn’t thought to look for. If you use R, there are plenty of snippets of code that make a summary plot of data, with frequency distributions and Q-Q plots. So give it a go. Work with your data from the bottom up – you never know what you might find.

P.S. Happy New Year everyone – There’s a surprise coming for Neuromancy next year…

Is science turning authoritarian?

Neuromancy — Sun, 05 Dec 2010 00:47:43 +0000

Going through the things I’d bookmarked to write about, I found an article in The American entitled “Science turns authoritarian”, in which it is argued that science has let itself become co-opted by politics, has begun to espouse view points rather than describe the world, and that as a result science is losing its credibility.To support this view, the following graph of frequency of so-called “authoritarian” phrases is presented.

I think two points are worth making about this, one of methodology and one of interpretation. Firstly, the graph is based on the frequency of these phrases in science reporting, rather than from scientists themselves. Whilst scientists could also be expressing similar attitudes, the position that these data support is that science is perceived to be to more authoritarian because those who report on it are perceiving it to be more authoritarian. Also, although the article suggests that if the increase were just a result of an increase in print volume then the frequency of all the words would increase equally. I’d argue that even if there was an increase in authoritarian tone, some of the phrases are going to be more commonly used than others. I’d also venture that use of some of the phrases (especially “science requires”) aren’t necessarily authoritarian – “a successful career in science requires dedication”?

Secondly, I think that the implication in phrases such as “science says we must” is a conditional statement that we must do X if we want Y. Where Y is generally something like the continuation of life as we know it. The article even cites some examples e.g. “For example, science tells us we must reduce our global greenhouse gas emissions to prevent dangerous climate change.” You could read that as saying emissions must be reduced if we are to prevent climate change. Even if you read it as making the assumption that we should want to prevent dangerous climate change, it doesn’t seem like an unreasonable assumption.

Which leads on to a third point. I don’t think scientists are capable of, or should be entirely objective in the sense of robotic detachment from the implications of their work. If reasearch strongly suggests that if we don’t drastically change the way we are living things will get terminally hard for a lot of people, you’d have to be inhuman not to feel like it’s something that we should probably be doing. And it’s probably something we should be doing more of. I think it’s particularly telling when the article mentions that people don’t take kindly to being told what to do “particularly when they’re unable to evaluate the information that supposedly requires them to give up their SUV, their celebratory cigar, or their chicken nuggets”. In circumstances where the information clearly requires a course of action, but only those immersed in the field have the ability to evaluate the information, it seems almost reckless to not give any context.

See The Authoritarianism Claim at Only In It For The Gold for a similar but more concise take on things.

Quickie – Considerations of in vivo electrophysiology

Neuromancy — Sat, 27 Nov 2010 22:13:12 +0000

I’m just reading a paper on the response properties of monoamine neurons, and it starts out with a few points of consideration for electrophysiological recording of neurons in vivo. I thought I’d relay them here.

The evidence that you are recording from a given type of neuron (dopaminergic, serotinergic etc.) is almost always indirect. It is usually based on electrophysiological characteristics determined by previous in vitro studies, using intracellular recording and neuron staining.
We can describe the general behaviour of a group of neurons which contain/release a particular neuron, as they often appear to behave in a homogenous manner. Jacobs notes “this may change, of course, as more detailed analyses are carried out.” The paper was published in 1986, and for my neuron of interest – dopaminergic neurons – this has indeed changed.
In order to properly understand the activity of a neuron, it has to be recorded from an unanesthetised animal, as beyond the advantage of being able to study behaviour and neural activity, anaesthesia can profoundly affect neuronal responses.

There’s also a nice little Q&A after the references, which seems quite quaint. It’s nice to read older papers to give a general overview of a subject before diving into the nuanced modern research, you just have to make sure it’s not so old as to be totally misguided and out of date.

JACOBS, B. (1986). Single Unit Activity of Brain Monoamine-Containing Neurons in Freely Moving Animals Annals of the New York Academy of Sciences, 473 (1 Neurochemical), 70-77 DOI: 10.1111/j.1749-6632.1986.tb23604.x

I’m learning about bees!

Neuromancy — Tue, 09 Nov 2010 18:44:34 +0000

Just on my way home from my second day as educational resource assistant for a project to encourage people to keep bees. My job involves creating educational materials for primary schools to accompany the project and highlight the plight of bees. This means I’ve spent the last couple of days reading the kind of thing that already goes out to schools, and reading lots of interesting things about bees.
Unfortunately, most of the stuff that I’m finding most interesting probably isn’t as relevant to the ecological importance of honey bees. Like how they don’t just have regular old dull compound eyes like I thought they did, but that they’ve also got three simple-lens ocelli, which are only really sensitive to light, and have been suggested to be part of some sort of orientation mechanism. Or that they don’t have centralised brains, but instead just have bunches of nerves down their body. Or that they can not only detect colour and patterns, but are capable of remembering them reasonably well several days after just one exposure, and almost perfectly after three! Or that their sting evolved from the ovipositor, but in some species it’s in a weird halfway stage where the sting is more like a sword.
The irreplacable role of bees in the economy, and more importantly the environment is probably easier to pitch as a classroom topic to primary school teachers, but I’m getting into it, and it seems a shame for it to go to waste.
So expect some bee related science over the coming few months. I’ll try and keep it neurosciencey…

Scatterplot matrices and MANOVA in R

Neuromancy — Wed, 03 Nov 2010 23:02:12 +0000

I’ve got a deadline to get one of my thesis chapters drafted by the end of the week. It was the end of last week, but all the unemplyment and job seeking stuff got in the way, and the analysis has been a little bit harder than I expected, so it got pushed back (oops, guess I should be used to that by now).

Anyway, I’ve been doing some of the analysis in R, and I thought I’d stick it up here to get some tips/feedback. Please bear with me on the state of the figures, they’re a bit sketchy at the moment.

Click ‘More’ for tedious technicalities!

The data are for 16 dopamine neurons and their responses to a light flash and trigeminal nuclear complex stimulation (mostly Sp5I). The variables are mean firing rate (Hz), average total spike width (ms), average spike width from spike onset to first trough (ms), response onset latency (s), response peak latency (s), response duration (s) and direction of the first component of the response (inhibited, unresponsive, excited).

Light flash data

	dafr	daapwidthtot	daapwidthtrough	daonsetlat	dapeaklat	dadur	group2
1	7.66	5.8	1.3	NA	NA	NA	Unresponsive
2	5.17	5.3	1.6	NA	NA	NA	Unresponsive
3	2.14	5.3	1.7	NA	NA	NA	Unresponsive
4	2.94	6.6	1.5	NA	NA	NA	Unresponsive
5	2.49	4.5	1.2	NA	NA	NA	Unresponsive
6	2.44	5.9	1.9	0.1	0.12	0.16	Excited
7	5.45	5.7	1.6	0.14	0.16	0.1	Excited
8	1.81	5.1	1.5	0.04	1	0.2	Excited
9	1.11	5.2	1.6	0.02	0.06	0.3	Excited
10	4.25	4.7	1.5	0.08	0.1	0.1	Excited
11	4.49	5.6	1.9	0.12	0.12	0.6	Excited
12	3.45	4.6	1.5	0.08	0.18	0.12	Excited
13	3.25	4.7	1.5	0.02	0.06	0.8	Excited
14	4.21	5.3	1.6	0.16	0.42	0.36	Excited
15	3.98	4.5	1.2	0.06	0.1	0.06	Excited
16	0.45	5.3	1.5	0.08	0.14	0.32	Excited

TNC stim data

	dafr	daapwidthtot	daapwidthtrough	daonsetlat	dapeaklat	dadur	group
1	5.17	5.3	1.6	0.02	0.04	0.08	Inhibited
2	1.11	5.2	1.6	0.06	0.04	0.08	Inhibited
3	2.94	6.6	1.5	0	0.04	0.18	Inhibited
4	4.49	5.6	1.9	0	0.06	0.24	Inhibited
5	3.45	4.6	1.5	0.04	0.04	0.1	Inhibited
6	2.49	4.5	1.2	0.02	0.02	0.06	Inhibited
7	7.66	5.8	1.3	0	0.02	0.04	Excited
8	2.44	5.9	1.9	0.02	0.08	0.18	Excited
9	5.45	5.7	1.6	0.06	0.08	0.08	Excited
10	1.81	5.1	1.5	0	0.06	0.02	Excited
11	4.21	5.3	1.6	0.06	0.14	0.38	Excited
12	3.98	4.5	1.2	0.02	0.02	0.04	Excited
13	0.45	5.3	1.5	0.08	0.14	0.32	Excited

I’ve managed to do a rough vis of the TNC stim data fine, with the scatterplotMatrix() (alternatively spm() for short) function from the car R package.

spm(trig[1:6], diagonal="qqplot", smooth=FALSE, by.groups=TRUE, group=trig$group, main="Trigeminal response/cell properties")

Producing a nice matrix, with linear regression lines for each group, and Q-Q plots on the diagonal, thus:

The problem comes when I try to plot the light flash data. Because the unresponsive cells (rows 1:5) don’t have values for response onset latency, peak latency or duration, I can only manage to make a plot of the dopamine cell properties for all 16 rows, and another plot for the cell properties and response properties for just the 11 responsive rows. Which I guess would be ok, but I’d like to plot all the rows and all the columns in just one figure, and have it plot all 16 points on the cell property panels, and just plot the 11 relevant points on the other panels. I’m not sure if that’s possible with the function I’m using though. It’d be nice to have the r^2 for the regression lines as well, but again, I’m not sure how.

The next step is learning how to do a MANOVA in R, but I spent all last night reading up on all the assumptions for all the different tests and now I’m not even sure whether the tests are appropriate…

Any help will be more than welcome!

A brave new world

Neuromancy — Fri, 29 Oct 2010 22:00:43 +0000

I can now blog from my phone! Hopefully this will mean more posts as I have the odd thought, probably of the short and badly spelled variety.

The staRt of a jouRney

Neuromancy — Thu, 28 Oct 2010 13:42:41 +0000

I thought I’d already done a proper introduction of my move into using the R language to analyse the data for my thesis, but I can’t find anything now… I’d previously been using GraphPad Prism, which was easy to use and everything was customisable at a click, but this meant it was a bit slow and bloated. MSExcel hasn’t even entered into the equation after the effort needed to display standard error bars, and also that some of the inference test calculations aren’t even right. Instead, I decided to try to teach myself R.

So here’s the start of my journey into R analysis, a “jouRney” if you will. R is an open source language developed from the commercial language S, which in turn derives from C. Straight out of the box it can do some reasonably advanced data extraction, analysis and visualisation. It’s real strength, however, comes from its openness. Being open source means that thousands of stat geeks around the world have developed add-on “packages” that develop R’s abilities into all sorts of areas of research.

So far, I’m just beginning to learn how to read data in, manipulate it and visualise it, but here’s the result of my first attempt to produce a heat map/contour plot of the firing rate of a dopamine neuron in response to a stimulus, across trials:

#reads in raw data from text file produced by heatplot.s2s
contourplot <- read.table("c:\\documents and settings\\craig\\my documents\\my dropbox\\R stuff\\rawoutput.txt", header=TRUE, sep="\t")
#trims off empty columns
contourplot <- contourplot[,1:40]
#sets table to contain only the data columns
contourplot <- contourplot[,c(2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40)]
#transposes the data
contourplot <- t(contourplot)
#makes table into data matrix
contourplot_matrix <-data.matrix(contourplot)
#Plots heatmap of matrix "contourplot_matrix"
filled.contour(x=seq(1, 300, length=20), y=seq(-0.5, 1.5, length=20), z=contourplot_matrix[1:20,1:20], nlevels=20, axes=TRUE, color.palette=topo.colors)

Which produces this:

A comparison of contour plot in R (right) and the raw data from Spike2 - Conotur plot rotated 90 degrees anticlockwise

Which is essentially a grid of the mean firing rate in 100 ms by 15 trial bins. Hopefully, I should be able to use these data in a principal component analysis, and see if there are separate components which change over time at different rates.

Eventually, I’m aiming to do all of my statistics in R, so we’ll see how it goes.

Sunday Links #8

Neuromancy — Sun, 24 Oct 2010 11:00:22 +0000

A collection of links for your consideration:

Cat cognition! Anne Corwin at Existence is Wonderful tests the mental abilities of her cats, with full experimental method and videos!

I’ve had a drafted post on what optical illusions teach us about how our vision works knocking around for some time now, but in the meantime I encourage UK readers to go watch the recent episode of Horizon on what we think we see isn’t necessarily what is there, and how our traditional senses could be augmented and expanded (summary on MindHacks). It’s on BBC iPlayer, so it will disappear soon, but it’s probably available on your local popular video hosting website.

A little bit of inspirational blogging: Not Exactly Rocket Science shares the influences and experiences of 130 (and counting) science writers, including Ivan Oransky of EmbargoWatch, neuro-writers Carl Zimmer, Jonah Lehrer, physicists, archeologists, climatologists, biologists from molecular to behavioural, and everything in between.

Speaking of Carl Zimmer, he has a new e-book out – Brain Cuttings: Fifteen Journeys Through the Mind. Vaughan Bell provides a review at Mind Hacks.

Speaking of e-books and Mind Hacks (these links don’t write themselves you know) Tom Stafford has a new e-book – The Narrative Escape, which is, in the author’s words “a long essay about morality, psychology and stories.” Also in Italian, and coming soon in Portuguese!

I promise this isn’t the start of a foray into political blogging. FiveThirtyEight.com debunks the concept of poll momentum.

Although previous research has suggested that caffeine consumption doesn’t improve alertness in habitual users, two recent studies found evidence that a cup of joe improves vigilance and control of visual attention in regular drinkers, and that it enhances working memory in extroverts. Abstracts are available, but articles are behind a paywall.

Breaking my promise

Neuromancy — Sat, 23 Oct 2010 15:53:27 +0000

OK, so I’m probably not going to keep up with my promise. It’s a busy old time at the moment, lots going on. Trying to fit real life stuff around finishing off my thesis (which requires more finishing that I would have liked) and trying to find a job! It’s probably a little unreasonable of me to expect a department who has no knowledge of me to take me on as a research assistant/associate when I haven’t submitted my thesis, and even more of a stretch when I’m applying for places which would prefer a slightly different skill set (making cultures of neurons then doing in vitro electrophys), but I’m enthusiastic, more than willing to learn, and above all optomistic. I’ve applied for a few lab technician jobs too, which, if I get them, would provide some valuable experience in a slightly broader range of techniques. Plus, they pay the bills.

For some reason, I seem to be on my way to being a statistics geek. Maybe I’m just not happy with being a regular science geek… I’m trying to learn the R language to do some of my thesis analysis, and I’ve applied for some data entry/analyst jobs. I’m also on the Civil Service Fast Stream scheme for Assistant Statistican! Erk, it’s all a little bit scary. I’m down in London on the 11th of November for an assessment day, don’t really know what to expect.

So that’s it. Now that I’ve said I won’t be around, I’ll probably end up being around more than when I said I was going to try and make the effort to blog more…

Bye for now!