So you have a cool result.

Maybe you’ve been optimizing an experiment for a while and finally have it working. You do that first solid trial and see a cool effect! Hurray!! Now what? The urge to run and tell your advisor might be strong, but you can definitely anticipate some of their next questions. Most people will tell you not to get too excited until you repeat it, but let’s go over some specific steps you might take to validate your result.

1) Think back to your experiment and whether the conditions and procedure went according to plan. This information should be in notes you took while performing the experiment. Were your cells healthy to begin with? Was the timing of the experiment precise? Did something strange happen during the procedure? If something went wrong that might affect the outcome, repeat the experiment.

2) Have you included relevant positive and negative controls in your experiment? Did they work? REALLY don’t bother to show it to your PI if you have omitted controls or the controls don’t work. Neither you nor your advisor will be able to interpret the result without working controls. Importantly, come up with a theory for why the controls didn’t work, fix, and repeat. Don’t just do the exact same thing again hoping for a different result unless you have thought about this first or have reason to believe something might have gone wrong (see #1 above).

3) Have you quantified the data? Don’t go with your gut or trust your eyes! Most of the time we are not capable of distinguishing by eye and are often biased by what we expect the result to be if not blinded. If the result is as unambiguous as night and day (there vs gone, many vs none, completely in the wrong place etc), go ahead and do a happy dance. Your PI might still not be as excited as you hoped if you haven’t repeated it, but in my opinion, a very solid trial with working controls and a quantified outcome is worth discussing. If there MIGHT be an effect but it’s hard to say, start quantifying any properties that might be relevant, even if it’s not what you were after initially. Sometimes what you were looking for isn’t true but something else relevant is. Collect some descriptive statistics that might be useful. All this is to say, your experiment isn’t done until you have analyzed the data.

4) Now that you have analyzed the data and know the true result, you’re ready to do a biological replicate (repeat the same experiment with new samples/cells/trials). If you get something different the second time, it’s tempting to do the experiment again to break the tie. A high degree of variation from experiment to experiment might mean that you haven’t controlled for some critical variable or that the parameter you’re testing is not the predominant causative factor for the process of interest. It is worth re-thinking both the design and premise of your experiment. Definitely tell your PI if the result is repeatable. They are likely holding their breath and will be ecstatic.

5) Finally, if the result is repeatable, what does it mean? Your experiment is also not done until you have interpreted the result. Make a model of what you think is happening biologically with your new information. Now test and try to break the model using ORTHOGONAL experiments. This is key. You can do the experiment 1000 times, but it can’t give you the confidence of multiple orthogonal results (particularly if the effect size is small or there is some variation). If the results or orthogonal experiments are consistent with your initial finding, this is the home run. Your result is solid and you’ve found something new! Congratulations!! You and your PI can both do the happy dance! If the orthogonal results aren’t consistent, it’s time to revise the model and come up with a new set of experiments to test it. This isn’t a failure. Herein lies the real beauty of science: systematically getting ever closer to discovery!

Avasthi Lab Career Development Week 2017

Last year, we held a Career Development Week for the lab, in which I encouraged everyone to take a little extra time to focus on their career goals. For this exercise, I provide a list of career development activities that the lab might want to participate in and assign points for each item (they don’t know the point value until after they’ve selected activities). At the end of the week, we have a meeting to go over everyone’s activities, tally points, and give a prize to the winner. We just got done with this year’s Career Development Week, and I think it was my favorite lab meeting of the year! The lab was already in a jovial mood when I arrived, Halloween candy in hand. Despite the success of last year’s exercise, the lab composition is a bit different and I wasn’t sure if they would really buy in to the idea. I was pleasantly surprised! Each person really threw themselves into these activities, enjoyed themselves, and gained something valuable from the experience!

This year, I gave them the following list to choose from:

  • Join twitter (must be public account and have name, some photo, bio)
  • Follow 20 scientists or science organizations on Twitter (points given for every 20 followed)
  • Make a website (must include your research interests, CV/publications)
  • Update your website
  • Make a career timeline (must include 1yr, 5yr, 10yr goals with decreasing level of detail)
  • Review a preprint and email comments to the senior author
  • Give a 5 minute verbal presentation to a scientist in a different field and ask them to rate their level of understanding afterwards (1=didn’t get it, 5=very clear and understandable).
  • Give a 5 minute verbal presentation to a non-scientist and ask them to rate their level of understanding afterwards (1=didn’t get it, 5=very clear and understandable).
  • Learn a skill you don’t know but feel you should (must show evidence of progress or proof of proficiency). This could be an experimental technique, proficiency with some software, type of data/statistical analysis).
  • Identify one career development area where you feel you are weakest and make a plan to improve it systematically over the next year (first step of the plan must be completed within career development week).
  • Identify a career/learning/teaching opportunity you plan to take advantage of (not one I have already alerted you to).
  • Write a few sentences about something useful you learned or gained by focusing on career development this week (I will use this or excerpts for a blog post on the lab website)
  • Consider a mentoring network and list additional faculty you don’t already know that you plan to ask for feedback and advice (the goal is to expand your connections and potential letter writers as well as diversify the advice you receive).
  • Come up with your own career development activity (the group will decide how many points it is worth at the end of the week).

Here are some of the highlights:

Soumita Dutta, a postdoc in the lab, added her information to her existing Twitter profile (@dutta_soumita) and followed 63 new people. I was really pleased that just by following some relevant contacts, she started to see the utility and benefits of SciTwitter! She made a beautiful website, made a career timeline, and did a verbal presentation to a scientist outside our field. Importantly she signed up for several services and contacted several people that would help her in her goal of ultimately transitioning to an industry position.

Brittany Jack, a grad student that ran away with the victory and set the bar high last year, tried to take advantage of points given for following scientists on Twitter (@bmichellej87). She brought her follower count to over 1000 (I ultimately gave only 0.5pts for every 20 scientists/science organizations followed so no one could win the competition from this alone). She updated her website, made a career timeline, and gave verbal presentations to a classmate and a non-scientist friend. She also made a plan to start summarizing her findings/implications and collecting this so that it might give her a better big-picture view of her work. What a great idea! She came up with her own set of questions about work-life balance that we will include in next year’s career development week. Here’s what Brittany had to say about this week:

“By focusing on career development this week, I was alerted to what I was already doing to advance my career and how much more I could be doing to advance my career. The most valuable part of taking the time to focus on career development is that we are reminded what are goals are and why we are doing everything we are doing. Sometimes, the overall goal gets lost in the day to day activities and I am refreshed when I take the time to think about 10 years down the road. What does that look like? How can I make an impact now on what happens 10 years from now?”

Evan Craig, our newest research assistant, joined Twitter (@evanwalkercraig) and hilariously said he went for “quality over quantity” to follow 7 people. This included everyone in the room and apparently Steven Colbert, who he claims Twitter forced him to follow somehow. Evan made a website, a career timeline, gave verbal presentations, and identified areas of weakness and plans to overcome. While he wasn’t quite sure what he wanted when he joined the lab, I was excited to learn he has decided to apply to grad school. We now have a plan to make sure he’s a slam dunk candidate! He came up with his own activity, which was to do a mock interview with lab members and faculty in preparation for grad school interviews. Here’s what Evan said about this week:

“Career development week allowed me take a step back, and properly gauge where I stand, not just in my immediate scientific work, but in my equally important long-term career path. It allowed me to recall goals I had not contemplated for a long time, and make the proper adjustments to keep pursuing those goals, or change my path accordingly. This week begged questions like has my research triggered an unforeseen interest that I may want to pursue? Did I attend a presentation where I connected with speaker’s research or occupation? Career development week helped me synthesize these experiences to make sure my “career development” is growing along with my actual feelings and interests. And I’ve realized if you aren’t instructed to actively think about career development, odds are, you won’t!”

Shengping Huang, another postdoc in the lab, added information to his existing Twitter profile (@shengping_huang) and followed many new scientists and journals. He made a career timeline, gave verbal presentations, and made a plan to address some weaknesses he identified. He was also the only person to review a preprint and send comments to the author (an activity worth a ton of points). I was thrilled to see he selected this activity and did it for two preprints. Shengping rightly suggested that next year, we put a point cap for repetitive activities so no one could game the system. Shengping said this:

“During this career development week, I reviewed two preprints and emailed the comments (suggestions) to the corresponding author. It is quite useful for me to get the story behind the preprint via communication with authors. It also helps me to recognize other people in various fields. I love preprints!”

Kate Fee, is a student in the lab that is part of a 6-year medical program, was in class and couldn’t attend the meeting, but she sent over her information so she could still participate (we saved her some candy). Kate made a career timeline, made a plan to improve some of her perceived weaknesses, identified a new learning opportunity to shadow an MD over her holiday break, and gave a verbal presentation to a friend who has been texting her with questions ever since! Here’s what Kate had to say:

“It is always fun for me to think about my future and all of the options that lie ahead of me. That being said, the amount of options is somewhat daunting. I have learned from this career development activity that I have a great many interests, and the best way to figure out how to tailor those interest towards furthering my career would be to investigate them, and to reach out to professionals and gain knowledge from their personal experiences.”

After the point totals were in, the winner was…Evan! He edged out Brittany by a few points and won a Visa gift card. In case Brittany crushed everyone again, this year I had a second place prize (Starbucks gift card), so everyone didn’t think chasing the prize was futile. Brittany won 2nd place.

Seeing all of the effort everyone put into furthering their own career goals and hearing how they felt about this activity was unbelievably rewarding. The lab seemed as energized about it as I was and Career Development Week will definitely be a long-term Avasthi Lab ritual. I particularly liked that everyone decided to present their research to a non-expert and I think this helped them step back and see the level of detail that is appropriate to help even science audiences better understand their work. Arguably one of the most useful activities, the career timeline, was also completed by each person. I hoped that setting long-term goals would help them take directed action in the short term and was happy to see that they all felt this was useful. As last year, I encouraged everyone to keep these their goals and efforts in mind and continue to invest in their future a little at a time.

Based on our overwhelmingly positive experience, I strongly recommend putting a spotlight on career development outside of the usual activities. If you have a similar plan in your own lab or decide to try a structure like ours, let us know what you did and how it went!

Feeding your scientific soul.

One thing that shocked me when I started my faculty position was the explosion of opportunities. As a grad student and postdoc, I remember clawing for opportunities that seemed all too infrequent. Upon starting my faculty position, I suddenly had access to people and things I never before could have imagined. Speaking invitations, requests to participate in panels, requests to work with me, quicker responses to my own requests of other faculty, increased potential was shocking. I was (and remain) thirsty for these opportunities.

Another new development that went hand in hand with this was the number of people telling me to say no to almost everything. “Saying yes to something means implicitly saying no to other important things,” they warned. “Focus on the things that will directly impact tenure,” they advised. All of these things are true. Certainly too much unfocused activity can sacrifice my ultimate goals of continuing to have the privilege of doing science for a living and leaving my mark on the field. But, while this advice is important to keep things in balance, I feel it shouldn’t be the whole story.

Personally, I’ve chosen to prioritize opportunities that meet any of the below criteria:

  • Does the opportunity directly help my research program?
  • Does the opportunity strongly contribute to changing the scientific landscape to the one I want for myself and future generation of scientists?
  • Does the opportunity help me expand my scientific circle? This can both have an indirect benefit for my research program and give me a seat at the table for discussions with influencers/decision makers.
  • Does the opportunity feed my soul? (Will it inspire new ideas or feed my curiosity about an aspect of my science? Will it allow me to help someone who otherwise would not get help? Will it push me out of my comfort zone to change my perspective on a scientific question? Is it fun?)

In short, articles about tenure like this by Radhika Nagpal and this by Matt Might speak to me so much more than the “learn to say no” articles. Inspirational and motivational advice can be so much more powerful than the practical caveats. Broader and more proactive thinking about the huge positive impact we can have on science and society (if we choose to) can sustain us through the [many] inevitable tough times. I hope everyone can experience the richness of the scientific career and find ways to meet their scientific goals while saying a resounding YES!

Try a writing checklist

Scientific writing doesn’t come naturally to most of us and is a craft that is honed over a lifetime of practice. One thing that can make the process of writing and getting feedback from your PI less painful (for all parties) is making a writing checklist. This checklist can be used on early drafts to help you catch common mistakes, improve the baseline quality of your writing and help the process seem less arbitrary. While the composition of this list can vary wildly, I’ve assembled one below based on some of the things that come up frequently when I’m editing drafts from my lab.

  • Is your background pertinent? Rather than generic background for your research topic, is the background sufficient to help readers understand only the information that will be included in the rest of the text? Cut unrelated background. Add information that a scientist outside your field needs in order to understand the concepts you’ll discuss (be aware of your audience).
  • Is your text free of unnecessary jargon? Never use shorthand to discuss experiments, ideas or tools. Do your sentences make sense if you replace all jargon/shorthand with their definitions (i.e. are all parts of speech appropriate)? Minimize and only use standard acronyms.
  • Have you used appropriate segues? Don’t jump around from one idea to another without appropriate explanation or transition. Can you lead into an idea by providing rationale and segue to the next by providing interpretation?
  • Have you ordered your results appropriately? It’s tempting to write things in the order in which they have been done chronologically. That is not always/rarely the most clear way to present outcomes. Take a look at all the data in aggregate, make a narrative and order accordingly.
  • Is what you’ve written factually correct based on the evidence/data/literature? Use language that makes it clear what is a result and what is interpretation/conjecture.
  • Is the main point (or main 2-3 points) of your writing exceptionally clear? Can a reader easily summarize the essence of what you’ve written? If the ideas and concepts seem muddled to you, they surely will seem muddled to the reader.
  • Attention to detail: have you caught all typos? Have you capitalized, italicized and hyphenated correctly in gene/protein/mutant names? Have you run the text through spell check and a grammar checker (many of these are available online)? These may seem secondary to the science, but sloppiness on this front is impossible to overlook and will completely change how your writing is perceived.
  • Have you gotten feedback from peers (including more junior colleagues in your lab who should be able to understand if your writing is clear)? This also means you have to finish a draft in time to get feedback and make changes! Also, after some rounds of revision, set aside and come back to edit yourself with fresh eyes. Some mistakes are impossible to see yourself once you’ve been staring too long at the same text. You can even give this checklist to your peers evaluating your writing as a sort of rubric for them to confirm you have adhered to key principles. This can improve the quality of the feedback you receive.
  • Is your draft of sufficient quality to represent you and your thoughts/work? Do a brief thought exercise. Your PI is definitely your safety net and your last line of defense, but assume for a moment that your PI isn’t going to stop you from making big mistakes. What if your draft were to go out into the public with your name on it but without additional input from your PI? Take responsibility for your writing. Many issues can be resolved by just asking if what you’ve written is truly the best you can do. This point is made beautifully here: (h/t @greenescientist for this link).
  • Importantly, go through this checklist BEFORE sending what you’ve written to your PI. If you catch errors that anyone can catch, your PI can focus on giving you the help that only they can provide. And if you repeatedly get a few different comments from your PI on many drafts, add additional items to the checklist for future use (i.e. ask yourself the questions you already know your PI is going to ask).

Formalizing some of the editing process can help writing consistency and quality. I hope this can serve as a starting point to stimulate your thinking on what you’d like to include in your own writing checklist.


Developing an idea filter.

I saw this great tweet today from @Holz_lab:

It’s so important for students/postdocs to develop a filter of their own.  If you don’t immediately pursue the newest idea, you have an opportunity to decide which ideas are best and should be pursued first. For example, if you have a list of 15 experiments and chose to pursue the best 2-3, you’ll probably feel fairly confident that you’re not wasting your time. In contrast, if you have one idea and plow forward with that, it may not end up being the best use of your time.

Currently, most of my newish lab still tends to jump down the rabbit hole of each experiment I suggest and I do end up making tweaks (large and small) to the ideas or the prioritization. To avoid excessive and frustrating redirection, I try to avoid thinking out loud and hold back thoughts that are tangential. But thinking out loud is a very useful exercise and tangential thoughts might provide interesting new leads. A more open dialogue about good (and bad) ideas is important. This is the messy business of science and getting clear thinking to emerge from the mess is an important skill that should be cultivated.

PIs may differ in how they expect you deal with their suggestions. However, regardless of your PIs mentoring style, developing your own filter for important experiments will serve you well throughout your scientific career. Write down all the ideas. Reorder them to let the best ideas rise to the top. Be prepared to defend your thinking to your PI who may agree or try to convince you otherwise. Think about how every new suggestion fits into your existing prioritization scheme before touching hand to glove. This will boost your confidence in your chosen experiments, increase your independence and reduce your frustration with PIs who seem to always move the goalposts.

Shaking up lab meeting

For a little change of pace, this past week I replaced our regular lab meeting with a writing exercise. Nearly every person in the lab was thinking about or actively working on a manuscript, so this seemed like a great time to ask people to formally think about how to put together a paper. My plan was to start the meeting with an abstract exercise outlined in this fantastic preprint by Halbisen and Ralston. The preprint provides some artificial figures that can be arranged in order and used to write a manuscript. Rather than asking the lab to write an entire manuscript on the fake data, I asked them to take a few minutes to arrange the figures in a reasonable order and write down their answers. The outcome was so much better than I had hoped. After collecting the responses, I was able to announce that no two people arrived at the same ordering except for myself and the newest undergraduate in the lab! This certainly got everyone’s attention since the least experienced person in the lab (in a room of undergrads, grad students and postdocs) was the only one that arrived at the same solution I did. I told everyone not to jump to conclusions and suggested that maybe we would be able to convince the others of our view or vice versa. A hilarious discussion ensued in which each person vehemently argued their point until we ultimately all agreed upon a final ordering with a common rationale.

After wrestling with the abstract considerations for organizing data, I asked everyone to then think about their own papers and order their data/figures in a way that they thought would be sensible for a manuscript. Some of them previously had been unsure about this task when thinking about their papers, but were now confident and easily able to proceed. If we disagreed, we discussed and were again able to reach a consensus on each ordering. To my delight, this exercise inspired several people to think of great ideas for experiments they could/should do. This is precisely the reason I feel strongly about organizing papers long before all the data are collected, but it was fantastic for all of us to see the benefits of it in real time! I too had some totally new ideas for everyone’s papers after hashing it out with them.

Overall, this combination of an abstract followed by concrete writing exercise was an unqualified success! I think everyone learned something, became empowered to push forward on their manuscripts and had great new ideas for their projects. And importantly, we had a ton of fun!

Update: Some are asking how long the exercise took. We only did the first dataset in the preprint for maybe 10-15 minutes and argued about it for another 20-25 minutes. The entire meeting including going over their own data took just over 2 hours.

Thoughts on picking a rotation lab

It’s almost summer and a new class of graduate students will be starting soon (we usually get at least a few students that do summer rotations). This got me thinking about how students select rotation labs. I put together this list of considerations that I think might help students trying to make decisions on where to rotate.

1. Follow your research interests. If you are truly enthralled by a subject or research area, do a rotation! It might seem obvious, but don’t overthink it. Even if many other students are rotating in the same lab, your genuine enthusiasm will likely make you stand out as a great fit for the lab. Also, don’t agonize about whether the PI has 5-7 years of funding already secured for you (hint: unlikely). These are tough times but there are mechanisms through the graduate program and university to make sure there’s never a lapse in funding that holds back a student.

2. Look for a mentor that will support multiple career paths and create opportunities for you. As a brand new graduate student, it’s likely you’re not sure what you’re going to do yet with your degree. Make sure you pick a mentor that is willing to have career conversations with you often and help you find opportunities that will help you decide or explore your chosen direction.

3. Be fearless! This is a great opportunity to explore and try new things. If you know nothing about a field, it’s hard to know if you would enjoy it. If there’s a lab you’re curious about but have no background in, give it a try! I am always surprised by what piques my interest after learning more about it.

4. Follow your nose to find the right lab team/family for you. It’s not a guarantee that you will get along fantastically with all members of every lab but it’s important that the lab is a place you can imagine being happy for many years. A culture of mutual respect should first and foremost be demonstrated and promoted by the PI. Listen to your fellow students about the reputation of the PI to catch any major red flags, but remember that a poor fit for one student may be an excellent fit for another. You know yourself and if you think you might thrive in a specific environment, give it a shot!

5. Remember, your needs, abilities and confidence will change dramatically throughout grad school. I always tell students that the uncertainty they feel as a first year graduate student won’t be the same when they’re a 4th year. You will learn a great deal and undergo a massive transformation as a scientist throughout graduate school. Look for a PI that will grow with you and is willing to give you more flexibility and freedom as you need/want it.

6. If you’re on the fence about rotating (or even if not), ask to sit in on a lab meeting! I have invited a lot of students to sit in on lab meetings or journal clubs. It’s a great way to learn about the lab culture and see how the lab and PI interact.

7. Picking a rotation lab and ultimately a graduate lab is a big decision, but it’s not terminal. Most places have the option of doing more than the required number of rotations. Even if you join a lab, things go very badly and you need to switch labs, your graduate program will be there to support you. It happens more often than you might think for many different reasons, so don’t let stress and fear rob you of the fun of this great exploratory time in your career!



Every once in a while on New PI Slack we revisit the discussion about lab expectations and onboarding documents. Our lab has a private onboarding document (private because it contains a lot of specific information about our data servers/logistics etc), but I wanted a public expectations document so that current/future lab members could see what we value and see what we think is important for them to succeed. I’m posting it below but it will ultimately appear as a heading on the main menu of this site.

What you can expect from me:

  • You can expect me to do everything in my power to find and create opportunities for you that will help you achieve your career goals, whatever those goals may be.
  • You can expect me to push you out of your comfort zone towards greater independence.
  • You can expect me to meet with you regularly to discuss experiments and future plans.
  • You can expect me to help you interpret experimental results, help you decide on next steps and help you prioritize those next steps (but will strive to guide your thinking rather than do it for you).
  • You can expect me to help you prepare grants/papers and edit mature drafts of written work (with earliest drafts scanned by labmates).
  • You can expect me to go over your career development plan with you at least once per year to see if you are on track for your goals and help you overcome any challenges.
  • You can expect me to send you to one conference per year if possible (preferably more if we can afford it or if you get your own funding).

What I expect from you:

  • I expect you to conduct yourself with the utmost scientific integrity. Never manipulate data or plagiarize written work (even if it’s your own previous work).
  • I expect you to keep a written record of your experiments in the designated electronic laboratory notebook that is updated preferably as you work but no less frequently than once per week.
  • I expect you to attend all laboratory meetings. I expect you to tell me in advance if you cannot attend. Experiments should be planned around these events.
  • I expect students and postdocs to identify funding opportunities and apply for all relevant/eligible external funding (including travel grants). Let me know of all submissions you have planned for the next 12 months.
  • I expect you to be present in the lab for most of the hours between 10am-4pm as that is when most of the lab activities and collaborative discussions occur. The remainder of your work time can be whenever you wish and flexibility in schedule is a perk of the job. Please let me know of any full days missed in advance. Resist the urge to compare your work schedule to others in the lab as the expectations for each career stage and situation are different. Students and postdocs might choose to work more than the 40 hours expected for research assistants because of their additional responsibilities in writing/project design, formal requirements to move to the next career stage and the extra time needed for career development activities. Work hard but stay sane. There will be periods of very high intensity work to meet deadlines/goals and periods you just need to just rest, slow down and reflect (see my previous blog post).
  • I expect you to create/revise your career development plan once per year and tell me how I can help you in any ways I haven’t thought of.
  • I expect you to treat all of your lab mates with respect, regardless of their career stage. Each person is an important contributor to our team that we can learn from.

Paper writing 101: Building a narrative

I was thinking about putting together a paper writing guide for the lab. Just thinking formally about how I write a paper was a useful exercise. I have lots of ideas on how to deconstruct this process and I started to make a giant list, but one thing I found that sort of encapsulated much of the advice was building a narrative. Let me demonstrate. Below I’ve written a nonsensical summary of a paper.

Title: D+G is the mechanism for how A works.

Summary: Here’s some background on what we’re studying and why it’s important. This is the state of the field/what we know so far. What’s critical, unknown and the question we wanted to ask is A. We figured out the answer by doing clever experiment B and found C. We did multiple orthogonal experiments that confirm that C is indeed true. We think C being true means that the most likely explanation for what is going on biologically is D. There are some alternative explanations for what C implies if not D, but we ruled those out. If D is true, it would predict E and F. We tested E and F and found they are also true but the results are a little unexpected, suggesting that what’s really happening is D+G. D+G would predict that H is true. We tested H and it is indeed true. All our data point to D+G as the mechanism for how A works, finally answering this open question in the field.

My feeling is that many people when starting to write a paper for the first time, can’t see the forest for the trees. It’s easy to get stuck in the details when day to day, you’re thinking at the level of troubleshooting individual experiments. Writing a narrative of the paper will serve as the North star for your paper writing process.  The above is just a stupid example and can vary wildly. When starting a paper, you should know A, have key result C in hand and interpret it to have a proposed idea for D. In my opinion, you don’t really need much else to begin. When writing this narrative, it will become painfully obvious that if you want to make your argument, you’re missing some pieces. That will help you decide which key experiments to add. This is why you shouldn’t wait until you think you’re done collecting data to write a paper. You’ll also notice that I didn’t write the title until I finished the narrative (I never would have known what letters to include!). Again, this highlights the point that you don’t really know what your story is until you go through the iterative process of adding data to round out the narrative.

If you force yourself to build a narrative early and keep revising it, you should eventually get to the finish line on your paper.