All Categories

Estimation stats launches

9/6/2019

In 2018, we wrote a short editorial to introduce new users to estimation graphics. We first released this as a preprint, which was received well.

How best to display data from two groups? Our new @biorxivpreprint argues that the best replacement for t-tests is the Estimation Plot. Unlike older statistical graphics, an estimation plot explicitly visualizes the comparison. https://t.co/EyhYnlsIYd pic.twitter.com/OkJ3B6K4MM
— Adam Claridge-Chang (@adamcchang) July 26, 2018

We then submitted this short piece, with the software for consideration by Nature Methods. After some time, and addition of several features, this was accepted and published. The publication was also met with a warm reception by scientists.

The new answer to a 77-year-old problem in data analysis, published today in @naturemethods. Instead of significance tests, use estimation graphics. Our software suite DABEST makes it easy for everyone to visualize effect sizes.https://t.co/UzwXJ7EUC5 pic.twitter.com/VtxyY0xaRM
— Adam Claridge-Chang (@adamcchang) June 19, 2019

Several other statistics-reform advocates recommended the software for scientists to use in their data analysis. Lewis Halsey, who previously highlighted the volatility of p-values, noted that DABEST and the estimationstats web app are the best methods to produce publication-quality estimation graphics.

I would really encourage people to check out this website. Easy to use and produces excellent, fully informed between-conditions plots. Far better than standard bar graphs that hide most of the information about your data. #stats https://t.co/9mqM0T9yWA
— Lewis Halsey (@lewis_halsey) August 7, 2019

Most recently, the Editor-in-Chief of eNeuro, the Society for Neuroscience's open-access journal, announced that he was changing the journal's policy to encourage the use of estimation statistics. This was backed up with a great explanation of the why and how of estimation by Robert J. Calin-Jageman and Geoff Cumming. The Editor-in-Chief Christophe Bernard wrote a personal account of how he came to adopt estimation:

#eNeuro @eNeuroEiC announces new author guidelines for statistical inference https://t.co/8YJSzaSGPY
— SfN Journals (@SfNJournals) August 2, 2019

This is a very promising development. We hope that this policy change at eNeuro spreads further, as we believe journal policy, when paired with the right tools for analysts, is a powerful way to transform statistical practice.

1 Comment

Postdoctoral Fellowship in Synaptic Neuroscience

9/6/2019

0 Comments

Neuroscience and Behavioral Disorders Program
Duke-NUS Medical School
While the structure and function of the pre-synapse is relatively well-understood, the complex structure attached to the post-synaptic membrane—known as the post-synaptic density—remains largely mysterious. The Human Frontier Science Program has funded a three-year project to study post-synaptic structure and function. This ambitious project aims to understand the dynamic organization of the post-synaptic scaffold, using a range of methods including: genomics, transcriptomics, proximity labeling, mass spectrometry, Drosophila neurogenetics, protein modeling, X-ray crystallography, cryo-electron microscopy, and electron tomography.

The Claridge-Chang Lab at Duke-NUS Medical School is seeking to appoint a Postdoctoral Research Fellow with a strong background in molecular neuroscience to join this project. The collaborative project will involve interactions with a number of other groups, including the Copley, Hoelz, Manser, and Robinson labs. The Fellow will investigate the structure and function of the postsynaptic scaffold.

Duke-NUS Medical School in Singapore is a collaboration between Duke University and the National University of Singapore. It has a mission for training educating physicians, training biomedical scientists, and condcuting research. The Claridge-Chang Lab focuses on analysing essential brain functions with behavioural, genetic, anatomical and physiological methods, with an aim to discovering some of the basic components and mechanisms that underlie psychiatric and cognitive dysfunction. The laboratory uses neurogenetic methods combined with ongoing molecular, instrumentation and analytical tool development and application to understand the circuits and molecules that support learning. For additional information please read other pages on www.claridgechang.net.

Job Requirements
Candidates must have:

A high level of motivation and a keen interest in neuroscience or biochemistry.
A PhD or equivalent in a relevant field.
Demonstrated knowledge, skills and expertise in neurobiology or biochemistry of neurons (or, in exceptional cases, a related field), with potential to achieve research excellence.
Strong publication record or, if recent graduate, high-quality manuscripts.
Advanced computer skills for recording, analyzing and organizing data.
An ability to work closely with other scientists and an ability to quickly learn new techniques.
Excellent communication, teamwork and project management skills. The ability to work harmoniously with a diverse workforce.
Experience in scientific sourcing & procurement.
Laboratory experimental skills, including the relevant skill set to perform the following techniques: cell culture; transfection methods; molecular biology (subcloning methods including restriction digests, PCR and gel electrophoresis); biochemical analysis; bioinformatic data analysis.
Fluorescence microscopy, laser confocal microscopy, image analysis (ImageJ, Zeiss LSM Browser, PhotoShop, Metamorph and/or Matlab) and neuroanatomical analysis (e.g. Neurolucida, FIJI, NBLAST).

Competitive candidates will have a knowledge of other techniques including:

Experience with the use and programming of software for data analysis and statistics (including Python, and/or other statistical packages like R or Matlab).
Drosophila genetics.
Behavioral experiments and analysis.
Apparatus design & testing.
Real-time apparatus control methods including stimulus delivery and control.
Methods for analysis of DNA and protein sequence (BLAST, Vector NTI, etc.).
Protein analytical methods.
Use of genetically-encoded indicators (GFP and derivatives, luciferase).
Analysis of gene expression (Q-PCR, Northern blotting, in situ hybridization).
Fine-scale brain dissection; Immunohistochemistry.
Graphic design (e.g. Adobe Illustrator).
Electronics.
Electrophysiology.
Optogenetic physiology, two-photon microscopy, and related methods.
Training will be available in some or all of these techniques where needed and appropriate.

Job Description
You will work to conduct research activities, including (but not limited to): planning; organizing; conducting; and communicating research studies within the overall scope of one or more research projects. Your main duties would include—but are not limited—to the following:

To pursue one or more projects using the above described methods and others as necessary.
To learn new methodologies when necessary or useful.
Maintain experimental progress.
To effectively analyse your results: apply qualitative and quantitative research techniques which include accurate in-depth assessment, interpretation and evaluation of research outcomes, conceptualize new ideas and develop plans for independent research which could have a considerable influence and impact on the relevant field.
To strive to uphold an excellent standard of statistical graphics.
To participate in the education and training of other staff as appropriate.
To acquire and maintain knowledge of the literature relevant to the project.
To attend appropriate scientific seminars and meetings and to remain current with the broader field of neurogenetics.
To regularly communicate your results to the rest of the laboratory members.
To write and review research papers, presenting research outcomes and develop connections with local and international researchers for collaboration work.
To work with colleagues on collaborative projects.
To contribute to project management, provision of guidance to more junior researchers as well as undergraduate and graduate students, occasional educational/instructional activities.
To share responsibility for the housekeeping of the laboratory area, general maintenance tasks and fly stock maintenance.
To conduct scientific sourcing and procurement.
To share responsibility with others for laboratory safety.
To perform any other duties as may be required from time to time to ensure the efficient running of the laboratory.

Apply Now
Interested applicants are welcome to email a detailed resume, supporting documents and contact details for at least two referees to claridge-chang.adam@duke-nus.edu.sg. Please note that your cover letter should indicate the reference number (Ref no: PN/ACC/RF-HFSP/201906). Only short-listed candidates will be contacted.

0 Comments

Farhan moves to start his own lab

3/28/2019

1 Comment

In November 2018, we celebrated Farhan's new job as an Assistant Professor at Hamad Bin Khalifa University in Qatar. While we were sad to see him leave, we are thrilled that he will be starting the next phase in his career. He is more than ready to take on the challenges of teaching and doing independent research.

1 Comment

Brainfest at ScienceCentre

6/7/2018

0 Comments

On the 19th, 20th and 27th of May 2018, members of the lab volunteered to help introduce Drosophila research to children at Singapore's Science Centre.

James, Joses, Sangyu, and Quyen spent three weekend afternoons interacting with kids showing them Drosophila flies and experiments in person. Hundreds of children had fun looking at flies!

0 Comments

A review of 30 years of memory genetics

5/8/2018

0 Comments

Four decades of memory genetics
We systematically reviewed over 30 years of Drosophila memory genetics. While replication is sparse, the replicated genes have high reproducibility. There is currently almost no evidence connecting genes into plasticity pathways.

Click to set custom HTML

0 Comments

Estimation Statistics App

4/1/2018

0 Comments

Everybody knows that significance testing and p-values are dying. But what will replace them? Our new web application https://t.co/kE5fIYlcZx provides user-friendly point-click estimation plotting for everyday data analysis. 1/2 pic.twitter.com/U3xVFzuKJ1
— Adam Claridge-Chang (@adamcchang) March 30, 2018

Over Easter weekend 2018 we launched estimationstats.com on Twitter. There was a lot of interest, and the information spread to scientists in many different fields. The upsurge in usage provoked some server memory outages. Requests included a version in R, and an option that can tolerate very large sample sizes (which the swarmplot cannot handle). We're happy that people seem to find it helpful, and welcome feedback on what doesn't work and what does.

0 Comments

Katarina Chlebikova, 1991–2018

3/30/2018

0 Comments

We mourn the loss of our colleague Katarina Chlebikova, a brilliant student and passionate scientist.

Katarina graduated with a degree in Biological Sciences from the University of Cambridge, doing her major project on learning and memory with Johan Alsiö. In the lab Jun - Sep 2013 she worked on Drosophila defense behaviors. Katarina went on to do her MSc at Edinburgh University before returning to Duke-NUS as a PhD student. She pursued her doctoral studies in the lab from Jan 2016. Tragically, Katarina passed away January 2018.

0 Comments

Why Google Docs is better than MS Word for Collaborative composition

8/24/2017

0 Comments

MS Word offers a huge range of advanced features. However, Google Docs is vastly superior for collaborative composition.

Simultaneous collaborative editing. Using realtime version control, multiple people can edit the same Google Doc—even the same sentence—all at once. Because you can see where someone else is working, you can easily avoid writing over the text they are working on. No version control system, not Word Track Changes, Merge Documents, nor even git–GitHub offer such seamless collaborative editing.
Threaded comments. In MS Word (at least the version I use on Mac), you cannot respond to a comment in a thread. All you can do is add a new comment, and your correspondent has to figure out which new comments are responses to their question. Threaded comments means that--rather than email back-and-forths--nearly all discussions about the text happen in context of the relevant text.
All-to-all communication. For MS Word, if one person is handling the version control by email and Track Changes approvals, they must be involved in every editing decision. This puts more work on them. In Google Docs, sections can be more easily delegated to subgroups in the writing team, who (due to thread comments) can easily communicate with each other directly.
Paperpile. If you use paid citation plugin Paperpile, anyone can insert a citation and anyone on the team can see the full citation. Paperpile is powerful citation software that is beautifully integrated into Chrome. See a paper on PubMed/ScienceDirect or a PDF? Click one button and it is now searchable and citable from within your GDoc.

Email and MS Word are technologies from 1971 and 1983, respectively. Their combination was never intended to be the final word in compositional version control. However, desktop word processors still do beat GDocs for typesetting; so I use GDocs for composition, followed by Apple Pages for typesetting the final manuscript.

0 Comments

Anion channelrhodopsins inhibit zebrafish behavior

8/13/2017

1 Comment

Zebrafish embryos move spontaneously. In fish expressing anion channelrhodopsin, green light induces paralysis.
Video by @GadistiAisha pic.twitter.com/GhgxKbbpG5
— Adam Claridge-Chang (@adamcchang) July 5, 2017

Check out our recent preprint, a collaboration with the Jesuthasan group. We find that anion channelrhodopsins are effective inhibitors of zebrafish embryos, and that GtACR2 is not trafficked to the plasma membrane well.

1 Comment

How to use custom fonts in Matplotlib

1/24/2017

4 Comments

After getting sick of the default font in matplotlib, and frankly annoyed by how difficult it seems to use custom fonts, Joses sat down and figured it out.

These instructions below are for Mac OS X, and assume you have installed Jupyter.

1. Install homebrew: go to and follow the instruction (no plural) here http://brew.sh/

2. Install font converter `fondu` with homebrew using the command:
brew install fondu
in Terminal. We will used `fondu` to convert the font of choice to 'ttf'.

3. Then find the font file you want to add. Open FontBook in your Utilities folder, find your font of choice, right-click on it, and select "Show in Finder".

4. Copy the font family to the folder "\Library\Fonts".

5. If this font already has the file extension `ttf`, you can skip to step 6.
If the font has file extension `dfont`, enter the following lines in Terminal:
cd \Library\Fonts
fondu <filename of font>
This will convert the .dfont file to a .ttf file.

6. Now we need to delete the font cache for matplotlib. In terminal, enter the lines
cd ~/.matplotlib
ls -l
to navigate to the matplotlib temp directory and list its contents.

7. There should be a file named either `fontList.cache` or `fontList.py3k.cache`. Use
rm <name of fontlist cache>
to delete it.

8. Now we need to edit the matplotlib parameters file. Start a new Python notebook in Jupyter and run the command
import matplotlib
matplotlib.matplotlib_fname()
in a new cell. This should give you something like:
'/Users/YourUserName/anaconda/lib/python3.5/site-packages/matplotlib/mpl-data/matplotlibrc’
which is the location of the parameters file.

9. Open this file in a text editor and scroll down to the Font section.

10. Find the line beginning with `font.sans-serif` and/or `font.serif`. Remove the '#' at the beginning of these lines.

11. Add the name of the font family to these lines. For example:
font.serif : Alegreya, Droid Serif, Bitstream Vera Serif, ...
font.sans-serif : Helvetica Neue, Source Sans Pro, Bitstream Vera Sans…
I have added Alegreya and Droid Serif to the serif family of fonts; Helvetica Neue and Source Sans Pro have been added to the sans-serif family. Matplotlib will look for fonts in running order, so put the fonts you want to use as default immediately after the colon.

11. Save the edited `matplotlibrc` file.

12. Restart your Jupyter notebook and run the following lines.
import matplotlib.pyplot as plt
plt.rcParams
and scroll to the Font section again to check that your settings have taken effect.

13. You can use the following matplotlib commands to change font family (sans-serif or serif) and the font. For example:
plt.rcParams['font.family'] = 'serif'
plt.rcParams['font.serif'] = 'Droid Serif’
You'll need to restart the notebook and run these lines again if you change the font.

14. To ensure that your edited parameters file will not be ignored when using seaborn, you need to import seaborn as
import seaborn.apionly as sns
This is only possible with the latest version of seaborn (v0.7.1).

-Joses Ho

4 Comments

<<Previous

Estimation stats launches

Postdoctoral Fellowship in Synaptic Neuroscience

Farhan moves to start his own lab

Brainfest at ScienceCentre

A review of 30 years of memory genetics

Estimation Statistics App

Katarina Chlebikova, 1991–2018

Why Google Docs is better than MS Word for Collaborative composition

Anion channelrhodopsins inhibit zebrafish behavior

How to use custom fonts in Matplotlib

Author

Archives

Categories