I'm a 5th year PhD candidate in Bioinformatics at a decent uni. I passed my quals and all I had left was to finish my project, write the dissertation and defend it. Unfortunately, a few days ago I discovered a major flaw in my project (involving a fast sequence search algorithm) which greatly limits its scope, to the point that it might not be publishable anymore/significant enough for the PhD. My funding runs out at the end of this year, so I don't have enough time to come up with something new for this part and build it (I still have another part to finish). I haven't yet told my PI about it (we meet this week) and it might still be possible to eke by with what I have, but there is a distinct possibility that I'll have to go ABD and graduate with just a Masters.
I started looking into open positions in Bioinfo (in NYC, where I'm based and from where I cannot really relocate) and it seems like most of them are looking for PhD level employees, although there are a few that will take MAs. My issue is that my PhD work has run somewhat counter to the mainstream Bioinfo field. It's a confusing blend of webdev (react + ts), probabilistic algorithms and machine learning. None of which I can boast particular expertise and none of which match the job requirements I've been looking at (which deal mostly with analyzing seq data).
Anyway, I'm not a software engineer but a lot of what I've done is programming-heavy and I'm wondering if at this point I should just start grinding leetcode and try to get a ML engineer or Webdev position in FAGMAN. Or if I'm just fucked and will basically have to start over doing entry-level bioinfo for near minimum wage? Anybody in the bioinfo field know what the job market is actually like right now?
Sorry for the blogpost Any Forumsentlemen. Just trying to figure out how I should move forward from here.
I'm sorry user. I've worked with a few bioinfo academic rejects that moved to webdev, so I know it's possible and they were very good.
They all seem to evoke your sentiment tho, it's very hard to get a job in bioinfo in the industry (even with a phd)
William Nguyen
i guess ur problem is too real for Any Forums folks. Just try to not expect something useful coming from here
Jordan Sullivan
This. The concentration of bioinf jobs at select facilities makes both the resulting work environments and office politics a fucking nightmare. Everyone there knows it’s far more specialised and thus at the mercy whim of the heads of those places,which they know (so the narcissists in those head positions absolutely fuck with their staff to keep them at bay from their own hard to find job).
If you move sideways into a broader software dev field (such as web dev) you’ll at least escape that trap. I’ve seen bioinf staff with PhD living nightmare work positions with low pay and exploitative work environments, and I can say to run quickly to an alternative. It’s too specialised.
Sebastian Hill
Incorrect. Some of us from the software dev space have worked in that field, to know how fucked that field is compared to other software jobs.
Tyler Jenkins
Thanks guys, this is helpful and sort of confirms my suspicions wrt to the work environment in bioinfo. It seems like you guys are recommending webdev over ML, is that just because it's easier to get a job in that direction?
I have experience with ts and react but I don't have any experience working in a dev team or with software engineering/project management. My webdev knowledge is also very haphazard and full of gaps and I don't have any projects other than my research. Should I just focus on trying to build some fullstack react projects (following youtube tutorials or whatever)?
Henry Perry
>I'm a 5th year PhD candidate in Bioinformatics at a decent uni. I passed my quals and all I had left was to finish my project, write the dissertation and defend it. Unfortunately, a few days ago I discovered a major flaw in my project (involving a fast sequence search algorithm) which greatly limits its scope, to the point that it might not be publishable anymore/significant enough for the PhD. My funding runs out at the end of this year, so I don't have enough time to come up with something new for this part and build it (I still have another part to finish). I haven't yet told my PI about it (we meet this week) and it might still be possible to eke by with what I have, but there is a distinct possibility that I'll have to go ABD and graduate with just a Masters. > I started looking into open positions in Bioinfo (in NYC, where I'm based and from where I cannot really relocate) and it seems like most of them are looking for PhD level employees, although there are a few that will take MAs. My issue is that my PhD work has run somewhat counter to the mainstream Bioinfo field. It's a confusing blend of webdev (react + ts), probabilistic algorithms and machine learning. None of which I can boast particular expertise and none of which match the job requirements I've been looking at (which deal mostly with analyzing seq data). > Anyway, I'm not a software engineer but a lot of what I've done is programming-heavy and I'm wondering if at this point I should just start grinding leetcode and try to get a ML engineer or Webdev position in FAGMAN. Or if I'm just fucked and will basically have to start over doing entry-level bioinfo for near minimum wage? Anybody in the bioinfo field know what the job market is actually like right now? > Sorry for the blogpost Any Forumsentlemen. Just trying to figure out how I should move forward from here.
You have the advantage of likely higher IQ to deal with the absolute chaos of web dev code, mainly because it’s often an abortion produced by teams of average programmers from disparate backgrounds other than university training.
The big issue with web dev is having the ability to understand spaghetti code without going into autism mode “muh formatted code” instead of fixing actual problems when working on other team members code.
The other issue is semi outsourced teams where you work on code written by foreign workers of such poor software quality without getting infuriated. This is what stops many devs who typically are highly focused on perfectly ordered code and rigorous quality processes.
This is one of the key factors that makes webdev unappealing to your standard software developer used to strict Asperger-tier development practises. Your intelligence might bypass this issue, provided you haven’t gone the “muh formatted code” cope argued by less than capable devs.
Austin Taylor
Working with spaghetti code is common in academia since most researchers aren't very good at coding, so I don't think I'll have any big issues there.
I'm currently trying to figure out what the best intersection of skills and requirements is so I can minimize the amount I have to learn. I have a solid stats and probability background that I kind of want to take advantage of, but looking at ML jobs they all want senior level engineers. Data science also has some overlap but I'd have to (re)learn some database tech like SQL, which I haven't touched almost a decade.
Nolan Clark
What exactly is the problem with your project? Are you using FASTA algorithms or what?
Nolan Morales
I'm not part of webdev or ML, but I have worked with ML people since I'm part of the data team who provides them with the tables and shit they can query for data.
If you want ML roles it's probably going to be "data scientist" roles. Most of which seems to be using python and applying whatever ML algorithm to predict something and make pretty looking graphs for executives on power BI or some other business intelligence software.
Whatever you do, as long as you go into software dev please get comfortable with writing automated tests and using git. Then just doing 1 leetcode a day should help you learn fundamental data structures that you may use for different types of problems.
Getting an entry level job in software is the toughest part though user, good luck on it.
Bit difficult to explain in so short a space, but basically it's an algorithm that can do fast (but approximate) sequence alignment (fast and memory efficient enough to run on a browser). I pitched it as being able handle queries that have at least 0.8 sequence identity (i.e at least 80% similar), but due the aforementioned flaw/mistake, it turns out it can actually only handle 95% similarity or higher, a four-fold decrease that greatly limits its scope.
>Whatever you do, as long as you go into software dev please get comfortable with writing automated tests and using git. I don't use CI but I'm familiar with jest and of course I use git and github for pretty much all my code (not always github, but certainly git--it would be like playing an rpg without save files otherwise). >Getting an entry level job in software is the toughest part though user, good luck on it. Yeah, I know someone that recently went through the hell that is entry level se job hunting and it's what I'm dreading the most. That and the fact that I just wasted 5 years of my life trying to get this PhD, only to (probably) fail at the end.
Wyatt Phillips
>If you want ML roles it's probably going to be "data scientist" roles. Most of which seems to be using python and applying whatever ML algorithm to predict something and make pretty looking graphs for executives on power BI or some other business intelligence software. Also, to add on, I'm literally teaching people how to do this as a GSI for a ML class this year so its definitely up my alley. I guess I just need to relearn SQL and start leetcoding.
Josiah Ward
>That and the fact that I just wasted 5 years of my life trying to get this PhD, only to (probably) fail at the end. Don't let it get you down. I wasted 10 years on a PhD in chemical engineering that I sidestepped into software development as a leaping point, due to a poorly scoped PhD topic. I ended up getting masters from it. Lost time happens.
Joe Vs The Volcano perfectly sums it up as "Follow the Crooked Path". The Straight Path is simple a fast trap to mundanity and a souless existence.
Gavin Gonzalez
Go back to facebook you blog posting faggot
Zachary Long
ahh ... as always, /g is /g
Brayden Sullivan
>10 years Jesus man, I feel your pain and I hope you're doing much better in your current work.
Josiah Robinson
Yeah. Things work out. The last 4 years were me working full time in software dev WHILE finishing the PhD. It is possible, just you have to put aside distractions like women etc.
Dylan Long
>I'm wondering if at this point I should just start grinding leetcode and try to get a ML engineer or Webdev position in FAGMAN Just do this. You have a PhD, you can memorize the leetcode shit and own the Indian coders that propagated this garbage interview style and snag that 300k job. Stop wasting time on shit like research, do something meaningful with your life (after 10 years of meaningless garbage at FAGMAN, you can retire and basically do whatever you want)
Carson Allen
Any chance you can still finish the PhD? Maybe take out a loan to cover expenses? While an ABD is ok it really is just so close to getting a PhD it would be shame not to go for it now. I guarantee if you leave now that you will never ever go back to finish it.
Jacob Robinson
>spend 10 years getting a phd in infosec. Become devops janny
>spend 10 years getting math phd in algebraic topology Become webdev
>spend 10 years getting a computational physics phd Become a data janny.
>Stop wasting time on shit like research Based. Been there. Research/academia is an appeal to the vanity of people as a trade off for not getting money.
Don't be tricked. The vanity of research/academia distills its practitioners into a group of toxic narcisstists of the highest psychopathy. It brings out the worst in people, not the PR-projected image of high-mindedness.
I have seen professors behave like the worst of rabid beasts. Consumed with their own grandiosity, they extinguish their humanity with the bestial pride of Nebuchanezar described in the bible.
"Brother; drive out power in yourself. Never let it fascinate you... " – Nestor Makhno
Eli Powell
Was thinking the same. Even the finding that an algorithm was not as good as predicted is an original finding in itself. Science isnt about discovering what is expected. Though I know in the real world "science" rarely adheres to its own so called high standards.
Lincoln Campbell
>Become devops janny Ignore this OP. Never fall for the devops meme. Its a trap. Been there. Done that.
Carson King
Thanks for the extremely helpful contribution you faggot. YWNBAW
Nathan Barnes
As someone working in academia and having supervised a few PhD and Master students I highly suggest talking to your advisor first. Usually the understanding is that if a PhD student fails its your fault, but if a postdoc fails its their fault. After all you cannot expect a PhD student to not make mistakes and you should within reasonable limits check on their work. Just finish what you got and let your supervisor figure out the rest. He/she should have a good idea of how to restructure your thesis and maybe highlight different parts. Usually your thesis committee will be determined by your supervisor and he may explain the situation beforehand. If the rest of your work is good/high quality and you can defend it well you still have a chance. I recently had a write a major correction for one of my papers because I had made a mistake which passed the reviewers as well. Luckily the majority of the results held true, but still very embarassing. Get back to me if you have any other questions.
