preLights, The Company of Biologist’s new preprint highlighting service, has now been running for more than three months. At the heart of preLights is the community of early-career researchers who select and highlight interesting preprints in various fields.
The preLights banner features cultured rat hippocampal neurons from Christophe Leterrier
We are now ready to grow our team of preLighters and are seeking early-career researchers, who are passionate about preprints and enjoy writing and communicating science. We welcome scientists across the biological sciences and especially those with expertise in Neuroscience, Bioinformatics, Microbiology, Ecology, Biophysics, and Systems Biology.
To join our team of preLighters, please send your application to prelights@biologists.com by the 30th June, 2018. In your application, please provide:
A short biography, telling us who you are and what you work on
A few sentences about why you are interested in joining our community
A preLight post highlighting a preprint of your choice
We have a flexible format for preLights, but your post should aim to include:
A short ‘tweetable’ summary of the preprint; background of the preprint; key findings of the preprint; what you like about this preprint; future directions and questions for the authors.
The post should reflect your personal opinion on the research in the preprint that you selected. Please also provide the URL link of the preprint. Your post should not exceed 1000 words.
To learn more about the ideas behind preLights, please read this introduction, or check out the interviews with current preLighters on their experience on our News page.
The current preLighter community
What’s in it for me?
This is a great opportunity for you to gain experience in science writing. You will get editing feedback from us and your peers and we aim to raise your profile as a trusted preprint selector and commentator. You will grow your professional network, and we are happy to support you by offering recommendation letters or in other ways.
But there is also a commitment; we expect you to select and highlight a preprint every one-or-two months.
We might not be able to accept all applicants, but are looking forward to welcoming our new preLighters.
You can find our recently published Nature paper here
Our story began two decades ago when my mentor, Klaus H. Kaestner, identified and cloned the transcription factor FOXL1, as being expressed in the mesenchyme of the mouse fetal gut (Kaestner et al. 1997). The position of FOXL1+ mesenchymal cells in such close proximity to the developing epithelium, as evidenced from In-Situ hybridization studies, suggested that FOXL1-expressing cells might be prime candidates to be key signal givers to the developing and adult gastrointestinal tract.
The early studies focused on the role of FOXL1 itself in regulating the intestinal epithelia by deriving mice homozygous for a FOXL1 null allele (FOXL1 -/-) (Perreault et al. 2001; Katz et al. 2004). FOXL1 null mice exhibit increased epithelial proliferation along with increased activation of the Wnt/β-catenin pathway, linking FOXL1 to Wnt/β-catenin pathway regulation.
The breakthrough came with a change in thinking. Klaus first realized that in order to study the role of FOXL1+ CELLS we should employ FOXL1 as a marker to trace these cells and genetically ablate them. The Kaestner lab derived two mouse models to kill FOXL1+ cells through the use of diphtheria toxin administration; FOXL1hDTR BAC transgenic mice that express the human diphtheria toxin receptor from a 170kb bacterial artificial chromosome that harbors all regulatory elements to direct transgene expression in subepithelial telocytes and Rosa-iDTR mice generated by crossing FOXL1Cre mice to a strain that produces the diphtheria toxin receptor from the ubiquitous Rosa26 locus in a Cre-dependent manner (Buch et al. 2005, Sackett et al. 2007).
Inducible ablation of FOXL1+ cells in adult mice caused a dramatic disruption to the intestinal epithelium, loss of stem and progenitor cell proliferation and the experimental mice died a few days after loss of telocytes had been initiated, demonstrating that FOXL1+ cells play a major role in stem cell function (Aoki et al. 2016).
I joined the Kaestner lab for my post-doctoral training during the time when these cell ablation models where being characterized in details. I am trained as a developmental biologist and therefore studying the potential cross talk between epithelia and mesenchyme was of great interest to me, even though at the time very little was known about the nature and function of FOXL1+ cells. In fact, we had no way to detect FOXL1 protein in tissue sections or biochemically, as multiple attempts at obtaining anti-FOXL1 antibodies by commercial outfits had been unsuccessful.
Thankfully, Chris V. E. Wright’s lab at Vanderbilt University came to our aid and generated multiple monospecific anti-FOXL1 antibodies for us. Antibody staining of mouse fetal gut showed a protein expression pattern for FOXL1 that was very similar to the one seen two decades earlier using radioactive In-situ hybridization to detect FOXL1-mRNA (Kaestner et al. 1997) (Figure 1 A-B).
However, the immunostaining for FOXL1 protein in the adult mouse intestine was disappointing at first. FOXL1 protein was present in the nuclei of selected mesenchymal cells; However, the abundance was very low (Figure 1C-D). On average, there were two to three FOXL1+ cells per crypt and their location was at mid-crypt region and along the villi core, in-addition to the crypt base where the stem cells reside.
Figure 1. FOXL1 marks a subset of mesenchymal cells during mouse development and in adult gastrointestinal tissue. (A-B) Mouse fetal gut E14.5 demonstrating by radioactive labeled probe (A) and immunofluorescence staining (B) nuclear FOXL1 mRNA (A) and protein (B) expression (red) in mesenchymal cells located in close apposition to the endoderm during development (outlined with immunofluorescence for EpCAM, green). (C-D) Sections of adult proximal jejunum, longitudinal section (C) and transverse section within the crypts region (D) showing FOXL1 expression localized to the nuclei of mesenchymal cells surrounding the crypt zone as well as alongside the villus core.
It is well known that the driving force for intestinal stem cell function is the Wnt/β-catenin pathway, which acts as a short range signaling-system. Potential cell types that might provide Wnt ligands should be in close contact with the stem cells. Do FOXL1+ cells touch stem cells? Do FOXL1+ cells express Wnts? Do they provide the essential Wnts that maintain stem cell identity? These were the key questions that I set out to answer.
With these questions in mind, I met Chris Wright for the first time in person, at the Gastrointestinal tract FASEB meeting in August 2015. During our little discussion, Chris mentioned “cytonemes”, cellular projections that are specialized for exchange of signaling molecules, and suggested that I investigate FOXL1+ cell structure. If FOXL1+ cells have long extensions and/or posses a unique cell structure, this might allow them to contact all epithelial cells.
Back in the lab, I had a clearer understanding as to which steps I should take –
I planned to inhibit all Wnt secretion from FOXL1+ cells and ask: Are stem cells affected?
I needed to sort GFP-labeled FOXL1+ cells (using a FOXL1-Cre ;Rosa-YFP mice) and prepare RNA seq libraries. Determine their gene expression profile: Do FOXL1+ express Wnts, and if so which ones?
I had to label FOXL1+ cells with a membrane reporter so that I could determine the extent of FOXL1+ cell structure
Targeting secretion of all 19 mammalial Wnt proteins can be done by deleting either Wntless or Porcupine, two essential Wnt processing enzymes, for which floxed mutant mice were already available. In order to inhibit Wnt secretion from adult and not fetal FOXL1+ cells, I need an inducible-FOXL1 driven Cre. With the help of fellow postdoc Kirk Wangensteen, I built a FOXL1-CreERT2 BAC and generated a new transgenic mouse line.
The next challenge arose when I tried to sort GFP-labeled FOXL1+ cells. To be able to sort FOXL1+ cells I had to selectively digest the mesenchyme from the intestinal epithelium in a single cell suspension. It was difficult to determine the optimal conditions to digest the mesenchyme as harsh digestion killed FOXL1+ cells, while mild digestion did not liberate any GFP-positive cells. Together with fellow postdoc Yue Wang, we devised a strategy to enable the selection to be successful. We isolated FOXL1+ cells, made RNA seq libraries and submitted them for sequencing.
The last piece of the puzzle was determinning FOXL1+ cell structure. To achieve this goal, I crossed our FOXL1 Cre mice to the Rosa-mTmG reporter in which FOXL1 promoter driven Cre activity lead to the expression of a membrane-bound version of GFP, which labeled the plasma membrane and allowed me to see the full extent of FOXL1+ cell.
The results were impressive; FOXL1+ GFP labeled cells were very large in extent and thus in contact with the entire epithelium from crypt base to the tip of the villi, with each and every single epithelial cell being touched by FOXL1+ mesenchymal cells!
During the same week I determined FOXL1+ cell structure, the RNAseq data was returned from sequencing. FOXL1+ cells indeed express a specific subset of Wnts and also the Wnt pathway inducers, R-Spondins. However, FOXL1+ cells also made high levels of Wnt inhibitors as well as BMPs, which are known to oppose Wnt signaling. How could this be, as we had shown through cell ablation that critical Wnt signals emanate from FOXL1+ cells?
I reasoned that since I had sorted FOXL1+ cells from anywhere along the crypt-villus axis for my RNAseq study, FOXL1+ cells might compartmentalize expression of signaling molecules based on their specific position along the crypt-villus axis.
To test this hypothesis, I contacted Shalev Itzkovitz from the Weizmann Institute, who had optimized a single molecule mRNA-Fluorescence In-Situ Hybridization (smFISH) technique for the mouse intestine. My goal was to focus on mRNA localization of different signaling molecules in FOXL1+ cell projections along the crypt-villus.
For this, I had to devise a way to label the full extent of FOXL1 cells, not just their nuclei. Unfortunately, I could not employ my Foxl1Cre Rosa-mTmG mice, as the reporter has a global tomato fluorescent protein expression, which bleeds through all analysis channels, thereby interfering with the smFISH signal. FOXL1 antibody staining would also not work, as it would label only the nuclei. What I needed was to identify a surface marker that could be specifically used to label the cells.
The RNAseq data revealed high expression of platelet derived growth factor receptor α (PDGFRα), member of the “villus cluster genes” characterized previously during gut development (Walton et al., 2012, Shyer et al., 2013, Shyer et al., 2015, Walton et al., 2016). Does PDGFRα label FOXL1+ cells? And if so, would it be possible to use it as a marker to label FOXL1+ cell extensions? YES! We demonstrated that all FOXL1+ cells are PDGFRα+.
In Shalev’s lab, Beáta Tóth combined PDGFRα immunostaining with smFISH and we were able to show regional differentiation in mRNA localization of different signaling molecules along FOXL1+ projections, with FOXL1+ cells near the crypt bottom producing abundant Wnt2b, a canonical Wnt pathway activator, while those further up the crypt-villus axis expressed high levels of Wnt pathway inhibitors.
I was puzzled by the unusual structure of FOXL1+ cells, which I also confirmed by electron microscopy. These cells are extremely thin but very large, with diameters in excess of 200 micrometer (for comparison, an intestinal epithelial cell is only about 10 micrometer in size). I was wondering if such unique stromal cells had been described in the literature, based on histological technologies.
Popescu named these cells “Telocytes” from the Greek words “telos” meaning end, “cytes” meaning cells. Telocytes are cells characterized by extremely long and thin projections called telopodes that may reach millimeters long and express PDGFRα in both human and mouse gut.
Apparently, neurons are not unique; telocytes also have long extensions that make direct contact with each other. Would it be possible to use the neuroscientists’ technology to study the 3D network of these cells? X-CLARITY is a method designed by neuroscientists for clearing tissues in order to visualize neurons in their 3D structure within the brain without the need for sectioning (Chung and Deisseroth 2013). Could we apply this technique to clear the intestine and visualize telocytes in their 3D structure?
In fact, clearing whole intestine and immunostaining for PDGFRα in green and EpCAM to label epithelial cells in red, allowed me to visualize the comprehensive stromal network of telocytes that form a plexus that supports the entire epithelium (Figure 2).
Figure 2. The 3D network of subepithelial telocytes in the adult intestine. Confocal imaging of cleared mouse whole small intestine using X-CLARITY. Immunofluorescence staining for PDGFRα (green) and EpCAM (red) showing the subepithelial network of telocytes.
My journey has just begun, and many exciting questions remain: Does FOXL1 label all telocytes? What is the origin of this remarkable cell? How and when do cells acquire telocyte characteristics? How do these cells compartmentalize signaling? What are the mechanisms by which telocytes signal to the epithelium.
A postdoctoral position is available in the laboratory of Dr. Lei Lei, in the Department of Cell and Developmental Biology, University of Michigan Medical School.
The research in the Lei lab focuses on cellular and molecular mechanisms underlying mammalian oogenesis. We are particularly interested in how the primordial follicle pool (i.e. ovarian reserve) forms during fetal ovarian development and is maintained in adult ovaries. The research in the Lei lab utilize broad cellular and molecular experimental approaches, including genetic mouse models, in vitro organ culture, live-imaging, single-cell lineage tracing, RNA-sequencing and proteomics.
The Lei Lab is seeking a highly motivated postdoctoral researcher to investigate the mechanism of intercellular bridge formation during fetal germ cell development, and the role of intercellular bridges in oocyte differentiation. The goal of this project is to identify novel fetal origins of adult ovarian diseases. Candidates must have a Ph.D. in biological science. Candidates should send a cover letter describing current research interests, a CV and the names and contact information of three references to Dr. Lei Lei (leile@med.umich.edu)
The Nerurkar Lab is looking for Postdoctoral Researchers with an interest in the interplay between molecular and mechanical aspects of vertebrate morphogenesis. Using the chick embryo, we combine live in vivo imaging, embryology and molecular genetics with engineering and physics approaches to study how developmental signals modulate physical forces that shape the embryo, and how forces in turn feedback on tissue growth and stem cell differentiation. Projects include early morphogenesis and patterning of the gut tube and brain, and organogenesis of the small intestine. Applicants must hold a PhD in molecular biology, development, bioengineering, or related field. Individuals with an embryology background who are interested in building an expertise in biophysical and quantitative approaches to development are particularly encouraged to apply.
Part of the Department of Biomedical Engineering at Columbia University, the Nerurkar Lab is located on the Morningside Heights campus of Columbia University in the City of New York. An academic reflection of New York’s excitement and creativity, Columbia offers a rich research environment, with boundless opportunities for collaboration with experts across engineering, biology, and clinical/translational disciplines. Interested applicants should contact Nandan Nerurkar at nln2113@columbia.edu.
Looking to conduct research in molecular biology and genetics? We are looking for a lab technician to assist in research on muscle stem cells, development, regeneration, disease, and evolution. More details about our research can be found at http://www.kardonlab.org/. Technician will assist in management of a mouse colony as well as conduct supervised research (leading to publications). Technician must be reliable, well organized, detail-oriented, excited about research and committed to working in our lab for at least two years. Prior lab experience is preferred (although not necessarily required), and class work in biology and enthusiasm for science is essential. Lab is located at the University of Utah in Salt Lake City, affording amazing opportunities for science and outdoor recreation. Looking for someone to start June-July 2018.
Please contact Gabrielle Kardon (gkardon@genetics.utah.edu) with CV, list of references, and a brief statement about why you are interested in the position. BS or BA required.
the GOEvol consortium proudly presents its 6th meeting #Sensation @GOEEvolution 2018 taking place in Göttingen from September 27th to 28th 2018.
The perception of environmental stimuli, their processing and integration is essential for any organism. Apart from the more familiar senses like hearing, seeing or tasting, there are sensory tasks performed by highly specialized animals, such as echolocation in bats or the perception of polarized light in insects. Sensory processing consequently also differs strongly between species. However, at the same time there are astonishing similarities between sensory modalities of phylogenetically distant animal groups, such as the shared cellular structure of light-sensitive organs or the genetic control and developmental origin of sensory cells. With methodological innovation, more and more species can be used for detailed analyses, which further expand the understanding of the evolution of sensation.
Because of the diversity of research and various methodologies in multiple (emerging) model organisms in the field of evolution of sensation we want to bring together scientists from a broad range of fields to reveal commonalities across disciplines.
Following the GOEvol tradition, we aim for an interdisciplinary symposium with an informal atmosphere with plenty of possibilities for social networking. If you enjoy small interactive meetings and the topic suits you, come along!
There are several slots for contributing talks and poster presentations. We strongly encourage interested students and researchers from all levels (Bachelor, Master, PhD and above) to register and apply for talks and poster presentations.
Moreover, we want to support parents to participate. Therefore, depending on the demand, we will be able to provide childcare as well as designated rooms for nursing.
Costs to register are 10€ for students, 20€ for Postdocs and PIs.
Invited speakers:
Sally Leys (University of Alberta, Canada)
Michael Bok (University of Bristol, UK)
Tobias Kaiser (MPI for Evolutionary Biology, Plön, Germany)
Robert Barton (University of Durham, UK)
Mirjam Knörnschild (Free University of Berlin, Germany)
Brigitte Schoenemann (University of Cologne, Germany)
Welcome to our monthly trawl for preprints in developmental biology (plus those hopefully relevant for developmental biologists).
May featured the usual catch of fascinating and beautiful work across the spectrum in the field, from Hox in mice and beetles, doublesex in beetles and bees, and three spinal cord regeneration preprints (incuding one using lampreys!). Our most prolific preprinter was Didier Stainier with four – a productive month for the Bad Nauheim-based biologist.
The preprints were hosted on bioRxiv, PeerJ, andarXiv. Use these links to get to the section you want:
Genome-Scale CRISPR Screening Identifies Novel Human Pluripotent Gene Networks
Robert J Ihry, Max R Salick, Daniel Ho, Marie Sondey, Sravya Kommineni, Steven Paula, Joe Raymond, Elizabeth Frias, Kathleen A Worringer, Carsten Russ, John Reece-Hoyes, Bob Altshuler, Ranjit Randhawa, Zinger Yang, Gregory McAllister, Gregory R Hoffman, Ricardo Dolmetsch, Ajamete Kaykas
Long-term expanding human airway organoids for disease modelling.
Norman Sachs, Domenique D. Zomer-van Ommen, Angelos Papaspyropoulos, Inha Heo, Lena Bottinger, Dymph Klay, Fleur Weeber, Guizela Huelsz-Prince, Nino Iakobachvili, Marco C. Viveen, Anna Lyubimova, Luc Teeven, Sepideh Derakhshan, Jeroen Korving, Harry Begthel, Kuldeep Kumawat, Emilio Ramos, Matthijs F.M. van Oosterhout, Eduardo P. Olimpio, Joep de Ligt, Krijn K. Dijkstra, Egbert F. Smit, Maarten van der Linden, Emile E. Voest, Coline H.M. van Moorsel, Cornelis K. van der Ent, Edwin Cuppen, Alexander van Oudenaarden, Frank E. Coenjaerts, Linde Meyaard, Louis J. Bont, Peter J. Peters, Sander J. Tans, Jeroen S. van Zon, Sylvia F. Boj, Robert G. Vries, Jeffrey M. Beekman, Hans Clevers
Pathogenic DDX3X mutations impair RNA metabolism and neurogenesis during fetal cortical development
Ashley L. Lennox, Ruiji Jiang, Lindsey Suit, Brieana Fregeau, Charles J. Sheehan, Kimberly A. Aldinger, Ching Moey, Iryna Lobach, Ghayda Mirzaa, Alexandra Afenjar, Dusica Babovic-Vuksanovic, Stéphane Bézieau, Patrick R. Blackburn, Jens Bunt, Lydie Burglen, Perrine Charles, Brian H.Y. Chung, Benjamin Cogné, Suzanne DeBrosse, Nataliya Di Donato, Laurence Faivre, Delphine Héron, A. Micheil Innes, Bertrand Isidor, Bethany L. Johnson-Kerner, Boris Keren, Amy Kimball, Eric W. Klee, Paul Kuentz, Sébastien Küry, Dominique Martin-Coignard, Cyril Mignot, Noriko Miyake, Caroline Nava, Mathilde Nizon, Diana Rodriguez, Lot Snijders Blok, Christel Thauvin, Julien Thevenon, Marie Vincent, Alban Ziegler, William Dobyns, Linda J. Richards, A. James Barkovich, Stephen N. Floor, Debra L. Silver, Elliott H. Sherr
The phylogenetically distinct early human embryo
Manvendra Singh, Thomas J Widmann, Jose L Cortes, Gerald G Schumann, Stephanie Wunderlich, Ulrich Martin, Jose L Garcia-Perez, Laurence D Hurst, Zsuzsanna Izsvak
Automating multimodal microscopy with NanoJ-Fluidics
Pedro Almada, Pedro Pereira, Siân Culley, Ghislaine Caillol, Fanny Boroni-Rueda, Christina L. Dix, Romain F. Laine, Guillaume Charras, Buzz Baum, Christophe Leterrier, Ricardo Henriques
NmeCas9 is an intrinsically high-fidelity genome editing platform
Nadia Amrani, Xin D. Gao, Pengpeng Liu, Alireza Edraki, Aamir Mir, Raed Ibraheim, Ankit Gupta, Kanae E. Sasaki, Tong Wu, Paul D. Donohoue, Alexander H. Settle, Alexandra M. Lied, Kyle McGovern, Chris K. Fuller, Peter Cameron, Thomas G. Fazzio, Lihua Julie Zhu, Scot A. Wolfe, Erik J. Sontheimer
C1 CAGE detects transcription start sites and enhancer activity at single-cell resolution
Tsukasa Kouno, Jonathan Moody, Andrew Kwon, Youtaro Shibayama, Sachi Kato, Yi Huang, Michael Böttcher, Efthymios Motakis, Mickaël Mendez, Jessica Severin, Joachim Luginbühl, Imad Abugessaisa, Akira Hasegawa, Satoshi Takizawa, Takahiro Arakawa, Masaaki Furuno, Naveen Ramalingam, Jay West, Harukazu Suzuki, Takeya Kasukawa, Timo Lassmann, Chung-Chau Hon, Erik Arner, Piero Carninci, Charles Plessy, Jay W Shin
Postdoctoral positions are available in the Parichy lab at University of Virginia. The lab studies development using zebrafish and related species. Current emphases include hormonal control over post-embryonic neural crest stem cells, plasticity in cell state, evolution of novel cell types, and mechanisms of pattern formation and cell–cell communication within zebrafish and across Danio species.
Methods include super-resolution time-lapse imaging on lab-dedicated microscopes, single cell RNA-sequencing, forward and reverse genetics, conditional manipulation of gene activity, and others.
The Parichy lab is situated in a highly interactive department with state-of-the-art facilities and UVA has outstanding current and historical strengths in morphogenesis and pattern formation. Training environment is excellent.
Applicants must have or be pursuing a Ph.D. and must have experience with modern methods of developmental biology. Prior experience with zebrafish is not required.
Applicants should submit the following to Dr. David Parichy (dparichy@virginia.edu):
• CV
• contact information for three references
• brief description of interests, experience and career goals
At the 2018 Spring Meeting in Warwick, the PhD and PostDoc representatives organises a series of events for young researchers. These ranged from asking questions such as “What is the secret to success?” through to some crafty synthetic biology and ending with the award of the 2018 BSDB writing competition.
A huge thanks to Michelle Ware and Alexandra Ashcroft for all their work in organising BSDB events for young researchers over the years. Learn more about the incoming PhD and PostDoc representatives of the BSDB here.
Career workshop
The career workshop kicked off the BSDB 70th birthday celebrations, with a topic focused on ‘building resilience and overcoming obstacles’. The idea was to utilise the knowledge and experience of the successful developmental biologists attending the meeting to give plenary and session talks. In total we had 18 tables, lead by Ottoline Leyser, Jordan Raff, Robb Krumlauf, Maria Leptin, Steve Wilson, Matthew Freeman, Fiona Watt, Magdalena Zernicka-Goetz, Jan Traas, Judith Kimble, Henrik Semb, Anne Ferguson-Smith, Patrick Lamaire, Jean Paul Vincent, Kate Storey, Austin Smith and Yohanns Belliache.
We wanted to have honest and open discussions about how to succeed in academia. ‘What is the secret to success, if there is any?!’. It is not very often you get to have opportunity to pick the brains of the leaders in your area of interests. This led to an afternoon of lively discussions.
Student/postdoc social
The format of the last couple of student/postdoc socials have been designed to get people to interact with people other than their lab mates! This year was no exception. This year’s activity was for randomly assigned teams to build models, within 1 hour, of any one of the following themes: fertilisation, plant development, somitogenesis, cell-fate determination and evo-devo.
We will let the pictures do the talking, but we were blown away by how amazing the models were and how much effort people put into making them. Megan Davey and Rita Sousa-Nunes judged the competition but had a hard time choosing winners since the models were excellent.
We are grateful to all the exhibitors who kindly donated prizes: Abcam, Class Learning, Proteintech, Philosophical Transactions B and Biographical Memoirs, The Company of Biologists, Stratech, 2BScientific and Indigo Scientific.
2018 BSDB Writing Competition
Finally, the student/postdoc events ended with the announcement of the 2018 Writing Competition. As part of the BSDB 70th anniversary celebrations, we initiated a writing competition where first prize was
a trip to the 77th Annual Society of Developmental Biology meeting (Portland, Oregon, USA). Katherine Brown, Aidan Maartens, Ottoline Leyser and Jonathan Slack judged the competition, selecting Daniyal Jafree as the winner.
So this year, is the final year for us organising these events. It has been a blast. Thank you all for making it such a wonderful and memorable experience. Please don’t forget to contact your future student (students@bsdb.org) or postdoc (postdocs@bsdb.org) reps for any comments or suggestions regarding the British developmental biology community.
See here for more pictures of the student/postdoc events.
The visualization of temporal data by line graphs has been documented and popularized by William Playfair in the 18th century (Aigner et al, 2011; Beniger and Robyn, 1978). Today, time-dependent changes are still depicted by line graphs and ideally accompanied by a measure of uncertainty (Marx, 2013). Below, I provide a ‘walk-through’ for generating such a plot with R/ggplot2 to visualize data from time-series. For convenience, example data and an R-script that performs all steps is available here. The data that I used is from Mastop et al (2017). After the make-over with ggplot2, the graph looks like this:
Some time ago, my favorite (commercial) software package for making graphs was no longer supported due to a system upgrade. So I was looking for a powerful and flexible alternative for data visualization. I ended up with R and ggplot2. First of all, the elegant data visualization delivered by the ggplot2 package is hard to beat. On top of that, it is completely free and it has a large user base. The biggest obstacle (in my experience) was getting used to the tidy data format that is needed as input. The conversion of ordinary, spreadsheet type data into long, tidy data is dealt with in a previous blog, and again briefly explained below. Before we start, we need to load two packages (these need to be installed first):
>require(tidyr)
>require(ggplot2)
The example data (available here) is read from a file in csv format into a dataframe (you need to set the “working directory” to match the location of the file):
The first column defines time and the other columns have data of individual cells corresponding to each of the timepoints. This data has a spreadsheet format, which is also named a ‘wide’ format. The first step is to convert it into a tidy data format, which is ‘long’. For details see my previous blog. The command to convert the df_wide dataframe into a tidy dataframe, df_tidy, is:
>df_tidy <- gather(df_wide, Cell, Ratio, -Time)
You can have a look at the first six rows of the dataframe to see what has changed:
>head(df_tidy)
Generating a basic graph
Once the data in the df_tidy dataframe is in long/tidy format, it can be used to generate a graph. The minimal instructions needed to display a graph are:
The dataframe that is used as input: df_tidy
The data used for the x-axis: Time
The data used for the y-axis: Ratio
To plot the data with minimal instructions use the function qplot():
>qplot(data=df_tidy, x=Time, y=Ratio)
Points/dots are used by default, but for these data lines are more appropriate:
The ‘qplot’ command provides a quick way to plot the data. To have a bit more control we turn to ggplot2(). The syntax is slightly more complex, but it allows to plot multiple layers. To reproduce what was done with qplot we need:
The aes() function is used for mapping “aesthetics”. The aesthetics specify how the variables from the dataframe are used to visualise those variables. In this case ‘Time’ is used for mapping the data onto the x-axis and ‘Ratio’ is used to map the data onto the y-axis. The geom_line() function specifies that the data is shown as a line. Inside geom_line(), aes(color=Cell) specifies that the line of each cell is mapped to a unique color. For more information on aesthetics and examples see this book chapter.
The geom_line() function defines that the data are displayed as lines. We can add another layer with a different geometric shape, geom_point(), to show the individual data as dots (in this example the dots will not have a color, since no aesthetic mapping is specified for geom_point):
The appearance of the plot depends on the sequence in which the layers are added. This is defined by the sequence in which the objects are added, i.e. the first object defines the first layer and next object is added on top. For instance, this code will generate a plot in which the lines are on top of the points:
The option of adding layers allows the addition of a summary of the data, for instance the average, or data for various types of error bars. To achieve this, a dataframe is defined to store the summary statistics:
We add the 95% confidence interval (95%CI) as a measure of uncertainty. Here we employ geom_ribbon() to draw a band that captures the 95%CI. To this end, we employ aes() inside geom_ribbon() to specify that the upper and lower limits of the confidence interval from df_summary define the borders of the ribbon. (Note that alternative methods that display uncertainty may be considered.)
The resulting figure only shows the data summary. Since it can be valuable to show the underlying data we will plot the data from individual cells as well. These data are present in another dataframe, yet it is possible to add these data to another layer. To demonstrate this, the original data from the individual measurements (from df_tidy) is added:
Improving the presentation and annotation of the graph
The graph shown above displays all the data of interest, but it looks cluttered. In the final session the lay-out and annotation will be changed to improve the visualization and communication of the results (see also ‘Graphics for Communication’).
To make the individual data less pronounced, the lines are plotted in grey. Also, the graph looks better if the ribbon of the 95%CI is drawn on top, so we move the grey lines to the first layer and plot the average and 95%CI on top of that:
To indicate where we performed a manipulation of the system another layer can be added. In this experiment a stimulus was added at t=44 and inhibited at t=146. To reflect this, we add a grey box with the function annotate():
The end results is a nice and clean visualization of the data, which is an improvement over the original visualization (made with commercial software, see insets of figure 6 from Mastop et al (2017)). The instruction to plot graphs with ggplot() usually consists of several different functions and may be daunting at first sight. I hope that providing this ‘walk-through’ that shows how to build a graph layer-by-layer lowers the barrier to start using R/ggplot2 for visualization of (temporal) data.
Acknowledgments: Thanks to Franka van der Linden, Eike Mahlandt and Jenny Olins for testing and debugging the code.
(1 votes)
Loading...
(No Ratings Yet)
Costs to register are 10€ for students, 20€ for Postdocs and PIs.
