Bytesize: Experiences in the nf-core Mentoring Program

Edit

0:01 (host) Hello, everyone, and welcome to this week’s bytesize. I’m very happy to have with me Alyssa Briggs and Mariana Guiladi. These were mentor and mentee, respectively, in the mentorship program that just closed. They will give us some of their impressions during the time there. Now it’s off to you.

0:34 Hello, everyone. Thank you for the introduction, and thank you for joining us here. I’m Alyssa. I will be going ahead and starting our presentation. Just to start with a little bit of background on us. We were both a part of the second round of this mentorship program. We’re just going to go over what were our assumptions before we started the program, what goals did we start with, why did we apply, what were we looking to get out of this, what did we see as some benefits or takeaways from the program, and then just advice for anyone interested, whether it’s applying as a mentor or a mentee.

1:23 I did just want to show this picture before we got started. As you can see, this program is quite global. It’s grown a lot since the first round and also is expanding into different regions as well. This is awesome. We’re reaching some more underrepresented communities. You can see me up here in the US and then Mariana up there in South America. This is a really cool aspect of the program, I thought. Whenever we come together for the larger group meetings in the program, it’s awesome to see what everyone is working on and the progress they’re making, all in different parts. Just a really fun aspect of the program.

2:03 More into what we’re going to talk about, the assumptions that we had before starting the program. One of my assumptions was that the mentor and mentee pairing was going to be based on experience level. I myself am fairly new with nf-core Nextflow. I’ve only been working with it for about a year. I was really hoping that I could help someone who was essentially completely new so that we could get the most out of the program as possible. What we found is this was definitely true. A lot of thought goes into making these pairings, not just based on experience level, but also on your research interests. Mariana and I both have some interest in viral or pathogen interactions with hosts. We were able to do some good work that we were both interested in because of that. This was a really great aspect about the program. Another assumption, this is one that Mariana specifically had, was that there would be strict guidelines and expectations. What we found is that the program is actually really flexible. It’s not strict. It’s not rigid. There are some expectations as to try to meet two hours a week, try to accomplish certain things without the program, but your options are huge. You can really cater that to your experience level, how much time you have to put into this, and your interests as well. That was great. Then another assumption was that you can learn Nextflow with no prior experience within the scope of the program. The program itself is going to just be about four months, which might sound a little bit daunting to try to learn everything about Nextflow in that time. But you absolutely can learn what you need to about Nextflow in that time period. You can go from having no experience, if that’s where you’re starting, to ending up with the skills needed to write and maintain a pipeline, if that’s what you’re looking to do.

3:58 We’re going to talk a little bit about our aims before the program. When I first applied, I didn’t have any experience with Nextflow. I was starting with bioinformatics. My main goals involved to learn the basics, how to run pipelines. But also, I had interest to learn how to write the code. Not only my own pipelines, but also help in pipelines of other people in the community. Alyssa is going to talk a little bit about her mentor site. So my goals when going into this program, I really wanted to contribute to our nf-core community in some way. I had a great experience with having a mentor of my own through my university, who really passed on the skills for nf-core and Nextflow to me. Having learned those, I really wanted to be able to pass them on myself. That was a big goal. Then I also wanted to gain experience and confidence by teaching the basics of Nextflow. As I’ve said many times, I was fairly new. I’ve done mentoring and teaching before, but never with anything related to bioinformatics. I definitely wanted to expand into that and just get more experience and confidence. I also went into this wanting to learn from my mentee and fellow mentors. Definitely, this isn’t a one-way street. If you’re a mentor, you’re still going to learn a lot yourself, and I definitely did. But that was an aim going into the program.

5:35 As Alyssa has said before, the program is very flexible. In our case, what we tried to do was weekly meetings. Each meeting, we tried to solve a problem. In the first meetings, we worked on the beginner tutorial of Nextflow. But we also later worked on pipelines that already existed. But we also tried to make sure that we were trying to implement new pipelines and new modules to contribute to the nf-core community. During this whole process, I was able to learn not only how to run these pipelines, but also how to write and test these new models and new pipelines that we tried to implement. Another experience that was really good was to see that not only we can contribute with coach the nf-core community, but also in other ways. For example, there is the Slack of the nf-core community. You can reach the questions inside of this Slack and answer and help people with their questions. Another way that we can help in the community is helping the translation of the training materials into another language than English, to help people around the globe to learn Nextflow. Some of my experiences. Again, like I said before, even if you’re a mentor, it’s not a one-way street. Make sure to learn from your mentees strengths. I think that was an awesome part of the program. Of course, I brought more of the nf-core Nextflow knowledge to the table. Mariana definitely had her own strengths as well, having been in bioinformatics longer than me. I learned a lot from her, which was a great experience.

7:27 Then something that’s general, just about working with anyone while you’re coding. It’s nice to hack through that code whenever you’re working on something and you hit a problem. You’re able to bounce ideas off each other and essentially solve that problem really quickly. You have these weekly meetings and they’re two hours. Some people might think that’s a big time commitment, but it’s really enjoyable. You’re working through what you’re passionate about and you’re having fun while doing it. Great experience. Then we chose two examples just to show of what can be done during this process. This is in our case, because Alyssa is a co-author of this pipeline for viral integration. One of the things that needs to be done to this pipeline was organize the code into subworkflows. Instead of the main pipeline having the bulge of the lines of code, we just call the subworkflows in the main pipeline. You organize this code into the new files to work the subworkflow. For me, it was an interesting experience to see how to make a pipeline more readable, let’s say, to see how pipelines are organized when we used Nextflow. This is one example. We try to show another example a little more complex, because three mentors, Sebastian, Kolja, and Alex, were working in a pipeline involving machine learning. They proposed to their mentee, Luria, to not only learn Nextflow using this pipeline, but also contribute with this pipeline. She was able to learn the basics, but also she made in the end of the program a contribution to the nf-core community.

9:33 We also want to give a little advice for future mentees and mentors. For the mentees, one thing that is good is try to learn Git and Bash before the program, at least the basics. Because since most of the codes for Nextflow and nf-core are on GitHub, this will help you with your first steps when you start the program. Also, if you already have some experience with other programming languages and you have a pipeline, try to implement this pipeline into a Nextflow pipeline. You’ll see that you’re going to have a good experience with it. We had some mentees that did that in our round, and they had good experiences doing that. Also, if you’re interested in running pipelines, try to use your own data or data that you are interested in. Because in this way, you will be able to see how to implement Nextflow in your everyday life in your career. My advice can apply both to mentors and mentees. My first advice that I pass on is: don’t be afraid to start small. Even if your goal in the end is you want to be working on pipelines, know how much experience you’re starting with and where you need to start. If that’s going through training documentation to start with or trying to just run an nf-core pipeline, don’t be afraid to take these smaller steps along the way. They’re worth just as much as being able to write a pipeline down the road. Like Mariana was saying, apply your own research interests. Look for scripts that you’re using every day. Try to turn them into modules for nf-core. You’re contributing to the community. You’re working on what you like, and you’re not necessarily starting so big that it becomes daunting. Going along with that, I suggest setting something like stepwise goals. If you’re starting from the beginning and you have to go through the training documentation and you’re trying to work modules and subworkflows and pipelines and all of that, make sure that you celebrate every success and every little thing that you accomplish. It can feel like a lot to get through, especially if you’re trying to write a pipeline. Just make sure you’re acknowledging all those little steps along the way. Then just don’t be afraid to ask for help whenever you’re unsure. There’s a whole community that’s going to be ready to help you. That’s going to happen in the regular Slack channels, whichever one is applying to the issue that you’re running into, as well as you’re going to have your own Slack channel just for your mentorship program. Even if you’re a mentor, you don’t have to know everything, feel free to reach out if you have a question. You’re going to learn something just as much as your mentee is doing.

12:28 Just a few messages for you to take and think about. The first one is, this program is not for people that already have experience or have no experience. It’s a program for everyone. Don’t be afraid to apply as a mentor or as a mentee because you’re going to have a good experience, not only learning Nextflow, but also sharing your knowledge with others. If you’re thinking about applying as a mentor and you’re insecure about you don’t have any experience or few experience, always remember that your experience is valuable for someone that doesn’t have experience at all with Nextflow. This person will be happy to be your mentee and have your knowledge. Last, as Alyssa said, always leverage each step on the way in the program because it’s a step closer to the goals that both mentors and mentees establish in the beginning of the program. All right. That being said, we’re pretty much done with talking about our experiences.

13:46 Round three of the mentorship is open. Applications are still open until May 15th. If you’re interested in applying as a mentor or mentee, make sure you get it in by then. This round is going to run, I believe, from June through the end of September. Then we also have linked here experiences in round two. That’s going to be the blog post that talks about the experiences and projects that the pairs worked on during round two. Thank you all for joining today. If you have any questions about our experiences, we’ll be happy to answer them.

14:20 (host) Thank you so much, Alyssa and Mariana. Now we are open for any questions from the audience. Maybe I can break the ice here. I have a question. No, Phil, you can come in. It’s fine.

(question) I have a question. Basically, there is one point that everyone always thinks about when they’re applying for something like this, which is the time commitment. It was mentioned before. I was wondering how much apart from these two hours that you spend each week together did you put in separately from these meetings, each of you in a way?

(answer) Yeah, so we had the two hour meetings. Then, of course, Mariana would work on some things in addition for a little bit longer and then whenever we needed to, we’d have conversations on Slack to get her through those issues. Mariana, how much time in addition did you spend? I think like four hours maybe, sometimes less, because we worked a lot together during these meetings. Most of the time when I was outside the meeting, I was focused on learning more about one thing that we discussed more than the time to solve a problem. But we had other pairs that didn’t have these weekly meetings. They just talked through chat on Slack. Then if they have a big problem to solve, then they have the meeting. It’s very flexible for both of the mentors and mentees to establish how many time they’re going to spend on these interactions.

(host) Thank you.

16:16 (question) Brilliant talk. Thank you. It was really, really nice to hear your experiences. We spent a lot of time planning how to build this and how formalized it should be and stuff. We’re never really sure if it’s going to work out. But hearing you guys talk about it really makes me feel very pleased that it was a useful experience for you both. I had a couple of just curiosity questions. I was wondering how much this overlaps with your day job, Mariana. Is this something you’re going to carry on using day to day going forward? And is it something you were already trying to do before the mentorship came along?

(answer) Yes. Right now, I am a biologist. I’m trying to transition into bioinformatics. For my PhD, I’m thinking about implementing a new pipeline that I will use. Translating Python and R scripts into Nextflow. My main goal is to try to use it all my everyday life from now on. One thing that I’m trying to do here with my peers is trying to show everyone that Nextflow is a great tool today to learn. I hope that I can convince others to use it also.

17:34 (question) We have a question in the chat. It goes, did you use any tools besides Slack, GitHub, and video calls to work together?

(answer) In the beginning, we used Gitpod, right? Yeah, we used Gitpod for a little while. Then we actually transitioned over to VS Code. We’re using Live Share a lot. That was super helpful in terms of trying to code together. I think those are the main programs we used.

18:07 (host) Maybe I should have mentioned that. Anyone can unmute themselves. I allow that now. If you have a question, just unmute yourself.

(question) Is there anything you would do differently if you were starting together again next week?

(answer) My main thing that I would do differently, since this was the first time I’d mentored nf-core Nextflow, anything like that, is to set, like I said earlier, more stepwise goals. Going in, we weren’t exactly sure what to work on, I guess, which was one of the harder parts of the program, is figuring out what to work on. Or if you want to convert some script into an nf-core module, trying to make sure that it fits the standards that nf-core has. Really working through that and trying to come up with those stepwise goals maybe in advance, I think, would have been helpful to give us a little bit more direction. That’s what I would have done differently. I totally agree, because in the beginning, probably I should have looked more into the pipelines and models that already existed. To gen an idea how to write on my own pipeline or something like that. But we had some good experience trying to find out which way to go. In the end, we had a good experience because we tried a lot of different ways. That was a good thing. But that’s what Alyssa said, I agree. That would be a good way to restart.

(comment) I guess that’s the downside of having it very flexible. When you say you work on anything, and then it’s difficult to know what to work on.

(speaker) Right, yeah.

20:11 (host) Do we have any more questions from the audience? Do we have any more questions from Phil?

(comment) I could probably think of more, but I’ve said quite a few already.

(host) In that case, I would very much want to thank you both for being here today and presenting the work also that you’ve done. I want to thank you also specifically for contributing to nf-core. It’s lovely to have you in the community. Of course, I want to thank the audience for listening in and the Chan Zuckerberg Initiative for funding the talks and as it happens, also this mentorship program. Thank you very much. Have a great day.