Sign in

Biplab Bose

@biplabbose.bsky.social
45 followers 114 following 71 posts

Biologist | Bookworm | Cycloholic | Birder | Trying to understand the design principles of Life using Mathematics and Physics | youtube.com/@sysbio fac.iitg.ac.in/biplabbose

PostsRepliesMedia
Biplab Bose @biplabbose.bsky.social · 12/01/2026
Classic floating-point quirk: 0.1 + 0.2 == 0.3 # false Julia: use ≈ instead. 0.1 + 0.2 ≈ 0.3 # true (Try \approx tab) #julia #computing
000
Biplab Bose @biplabbose.bsky.social · 01/01/2026
Assume gains are private, but losses are shared. What happens if this rule governs a system of interacting open subsystems exchanging resources with the environment? The system will have a sharp, switch-like state transition. Check our new pre-print arxiv.org/abs/2512.22839 #statmech #ComplexSystem
020
Biplab Bose @biplabbose.bsky.social · 20/12/2025
An increasing number of UG students appear to be struggling with mathematics. What we thought was a local, transient issue may actually be part of a wider, possibly global, problem. maa.org/math-values/... #math #engineering #edu
maa.org
Students are struggling. Why? – Mathematical Association of America
000
Biplab Bose @biplabbose.bsky.social · 10/12/2025
In Julia, function composition isn’t just notation… it's executable math > f(x) = x^2 + 1 > g = f ∘ sin > g(pi/2) # gives 2 #julialang #joyofcomputing #computing
000
Biplab Bose @biplabbose.bsky.social · 07/12/2025
A university is closing its Department of Statistics due to financial constraints. Really? In this age, where mathematics, computing, and AI/ML dominate every industry, closing a Statistics Department doesn’t seem like a good business decision. #justsaying nebraskaexaminer.com/2025/12/05/t...
nebraskaexaminer.com
‘This hurts’: NU regents back shuttering four academic programs at UNL • Nebraska Examiner
The University of Nebraska Board of Regents voted to eliminate four academic programs at the University of Nebraska-Lincoln.
000
Biplab Bose @biplabbose.bsky.social · 05/12/2025
To what extent can the foundations of thermodynamics & stat physics be built purely from combinatorics and probability—without relying on the usual concepts like work, heat, and energy? Are there any textbooks that develop the theory primarily from a probabilistic perspective only?
000
Biplab Bose @biplabbose.bsky.social · 04/12/2025
Don't be clueless! Better, understand how Julia thinks: @code_lowered 2 + 3 @code_lowered 2 + 3.0 @code_lowered 2.0 + 3.0 Code better. Code smarter. #JuliaLang #computing #sysbio #systemsbiology #qbio
010
Biplab Bose @biplabbose.bsky.social · 03/12/2025
High-throughput omics studies > lots of data transformations > 5,000 correlations to be tested through experiments. Toy models with few parameters > the design principles behind the complexity of Life. True Systems Biology: fewer spreadsheets, more ‘Aha!’ moments. #sysbio #mathbio #physicsoflife
000
Biplab Bose @biplabbose.bsky.social · 02/12/2025
Why burden memory? Take a sneaky peek and get the job done! A = rand(5,5) W = @view A[2:4,2:4] # take a peek, no copying @views W[1:2, 1:2] .= 0 # sneakily mutate A # check the mutation #julialang #computing
010
Biplab Bose @biplabbose.bsky.social · 01/12/2025
Explaining non-equilibrium systems to students: ‘Imagine a system that refuses to relax.' Students: ‘Like us?’ #justforfun #teaching #physicsoflife #physicsoflivingmatter
000
Biplab Bose @biplabbose.bsky.social · 30/11/2025
Schrödinger’s cat called. It is live and it wants a break. It didn’t sign up to explain superposition to every social media users. Enough!
010
Biplab Bose @biplabbose.bsky.social · 29/11/2025
Using ± in Julia code? Yes, because real models don’t come with exact parameters. using DifferentialEquations, MonteCarloMeasurements k = 1.0 ± 0.1 u0 = 1.0 ± 0 prob = ODEProblem((u,p,t)->-k*u, u0, (0,5)) sol = solve(prob) mean(sol[end]) #JuliaLang #computing
030
Biplab Bose @biplabbose.bsky.social · 27/11/2025
Temperature is not just a thermometer thing. Physicists, chemists, biologists, economists, and data analysts all use it differently. Maybe one day we'll find a single rule that links them all. Until then, temp keeps its cool by being complicated.
000
Biplab Bose @biplabbose.bsky.social · 25/11/2025
Another review article worth bookmarking: Signed Networks: theory, methods, and applications." arxiv.org/abs/2511.17247 Perfect for my Systems Biology course. #sysbio #complexsystem #graph #network #compbio #mathbio
arxiv.org
Signed Networks: theory, methods, and applications
Signed networks provide a principled framework for representing systems in which interactions are not merely present or absent but qualitatively distinct: friendly or antagonistic, supportive or confl...
010
Biplab Bose @biplabbose.bsky.social · 24/11/2025
Many bugs don't scream. They whisper in wrong units. Using units in your code can save you. Again, Julia helps in that. using Unitful D = 0.01u"mm^2/s" # 0.01 mm^2 s^-1 t = 100u"s" # 100 s √(2D*t) # 1.4142 mm #julialang #computing #sysbio #systemsbiology
020
Biplab Bose @biplabbose.bsky.social · 23/11/2025
Identical genome. Same environment. Totally different life choices. Turns out even cells have an identity crisis. #sysbio #physicsoflife
000
Biplab Bose @biplabbose.bsky.social · 22/11/2025
I get confused when philosophers and scientists debate “reality.” If I slap someone, they feel pain and may hit back. That’s real. So clearly, I’m missing something much deeper. #justsaying
000
Biplab Bose @biplabbose.bsky.social · 21/11/2025
In biology, we’re still busy naming T cells and debating EMT or autophagy — a bit like arguing over bird names while the real wonder lies in understanding how they fly. #complexsystem #PhysicsOfLife #sysbio
010
Biplab Bose @biplabbose.bsky.social · 20/11/2025
Dear developmental biologists, Julia comes with a Zygote that auto-differentiates. 😀 using Zygote f(x) = x^2 + 3x + sin(x) f'(2.0) # returns 6.58385 #Julialang #mathbio #qBio #compbio #sysbio #systemsbiology
120
Biplab Bose @biplabbose.bsky.social · 19/11/2025
At the beginning of this year, Active Matter researchers published a 'review of reviews' — yes, a meta-meta study. When your field grows so fast, you need a map of the maps. Worth bookmarking. link.springer.com/article/10.1... #activematter #physicsoflife #physicalbiology #complexity
link.springer.com
Metareview: a survey of active matter reviews - The European Physical Journal E
Abstract In the past years, the amount of research on active matter has grown extremely rapidly, a fact that is reflected in particular by the existence of more than 1000 reviews on this topic. Moreov...
010
Biplab Bose @biplabbose.bsky.social · 17/11/2025
Is it weird that I prefer old biology papers—the ones with two figures, zero rainbow plots, and actual ideas? Or am I just allergic to 100-page supplements that explain nothing?
000
Biplab Bose @biplabbose.bsky.social · 16/11/2025
Every time I get overwhelmed, I remember- a lot of statistical physics is basically fancy bookkeeping for balls in boxes.
000
Biplab Bose @biplabbose.bsky.social · 15/11/2025
When science books cost a fortune, only publishers win. Open sharing—GitHub, arXiv—helps every learner. Isn’t it time to free ourselves from the 'medieval' prestige of being ‘published by a reputed publisher’? #scicom #edu #doingsciencd
000
Biplab Bose @biplabbose.bsky.social · 14/11/2025
Wet lab? Dry lab? I prefer the Symbolic lab. Math-Bio isn't just HPC runs— derivations & proofs matter too. As always, Julia helps. using Symbolics @variables x y a b c d m n f = [a/(1+y^n)-b*x, c/(1+x^m)-d*y] J = Symbolics.jacobian(f,[x,y]) #Julialang #mathbio #computaionalbiology #sysbio
030
Biplab Bose @biplabbose.bsky.social · 13/11/2025
Calling any large, well-organised dataset "Systems Biology" is a bit like calling a census "social theory." Essential, powerful, and impressive — yes. But the explanations, through generalised theories/models, still have to be earned. #systemsbiology #sysbio #physicsoflife
000
Biplab Bose @biplabbose.bsky.social · 12/11/2025
PhD defense (Dramatized based on real events): You (examiner): That statistics & conclusion are wrong. Needs correction for repeated tests. Candidate: 🤔 (flustered). Supervisor: No issues! 😡 This work is already published. How often do you face such a situation? #worldofbiologists #bio
000
Biplab Bose @biplabbose.bsky.social · 11/11/2025
My physicist friends love universality — until the system starts voting, reproducing, or trading stocks 🤪 #complexsystem #physicsoflife #physicsoflivingmatter #systemsbiology
000
Biplab Bose @biplabbose.bsky.social · 10/11/2025
A = [1, 2, 3] B = [3, 4, 5] x = 9 x ∈ A ∪ B # -> false Writing code using symbols in Julia feels like scribbling in a math notebook — except this one actually runs. It takes some time to get habituated, but then you get hooked. #julialang #computing
071
Biplab Bose @biplabbose.bsky.social · 09/11/2025
Every time I try to post on #mathbio, I realize 300 characters can’t hold a nonlinear thought!
000
Biplab Bose @biplabbose.bsky.social · 08/11/2025
In the age of AlphaFold, should we still teach homology modeling, threading, and other classical structure prediction algorithms? In the era of generative AI, the hardest question in teaching isn’t what to include—it’s what to leave out. #teaching #education #computationalbiology #AI
010
Biplab Bose @biplabbose.bsky.social · 07/11/2025
With a channel capacity of one bit, cells have ruled the world for millions of years. Maybe we need a different information theory to understand them. #complexity #systemsbiology #complexsystems
010
Biplab Bose @biplabbose.bsky.social · 06/11/2025
PS Dodds cheekily reimagines the entire history of science as a three-volume saga. Beneath the wit lies a serious claim: complex systems aren't a fad but the long-awaited finale of scientific understanding. arxiv.org/abs/2511.00368 #complexity #systemsbiology
arxiv.org
The Three Books of Science
We venture that the long evolution of science may be viewed as unfolding over three blurred epochs. The first epoch spans the slow, haphazard, error-ridden realization of scientific truths along with ...
010
Biplab Bose @biplabbose.bsky.social · 05/11/2025
Robotic Experiment > Data > AI-driven [Model > Hypothesis > Experimental Design] > Robotic Experiment: This "closed-loop science" is set to become a reality soon, cutting costs & increasing efficiency. But data is not the limiting factor in Science; Our limited abstraction & generalization are.
010
Biplab Bose @biplabbose.bsky.social · 04/11/2025
Yes, current LLMs hallucinate and fail in many silly tasks. Current AI is surely not the same as human 'intelligence.' But why do we get hyperactive on social media with glee every time we find an error by an AI agent, LLM, or robot? Our existential crisis?
000
Biplab Bose @biplabbose.bsky.social · 03/11/2025
Julia keeps it rational. 1//3 + 1//6 == 1//2 Exact math. No floating-point drama. #JuliaLang #lovingJulia #sillyposts
021
Biplab Bose @biplabbose.bsky.social · 02/11/2025
Statistical physics was built to understand gases. Now we use it to understand flocks, tissues, and tumors. Same mathematics, new meanings—order out of interaction. Perhaps we should develop new thermodynamic variables beyond V, P, and T that are more suitable for living systems. #complexsystems
010
Biplab Bose @biplabbose.bsky.social · 01/11/2025
Between the animate and the inanimate lies a missing theory — not of matter, but of matter becoming intention. Or the concept of teleonomy is just another trap? Possibly! #complexity #physicsoflife
010
Biplab Bose @biplabbose.bsky.social · 31/10/2025
Physicists see the forest, mathematicians see the pattern, & biologists... name every tree, catalog leaf shapes, & write three papers on bark texture before asking why forests exist. Wish my friends from the Biology community would read and re-read the old classic: sciencedirect.com/science/arti...
sciencedirect.com
Can a biologist fix a radio?—Or, what I learned while studying apoptosis
000
Biplab Bose @biplabbose.bsky.social · 30/10/2025
If I want to teach undergraduate students about driven non-equilibrium systems, which two toy models should I choose? The mathematics must be intuitive, and the simulations must be easy. What do you suggest? #statmech #physics #complexsystem #systemsbiology
000
Biplab Bose @biplabbose.bsky.social · 29/10/2025
In #Julia, even the script evolves with time (without mutation) 😃 julia> @time sum(rand(10^7)) [0.086688 seconds, 57.81% gc time] julia> @time sum(rand(10^7)) [0.026221 seconds, 29.82% gc time] . . julia> @time sum(rand(10^7)) [0.017704 seconds, 2.34% gc time] #Julialang #computing
110
Biplab Bose @biplabbose.bsky.social · 28/10/2025
Gradient systems give you landscapes. Non-gradient systems give you the weather. That's why cellular fate feels more like climate chaos than a downhill stroll on Waddington's landscape. #systemsbiology #dynamicalsystems #complexsystems
000
Reposted by Biplab Bose
Ricard Solé @ricardsole.bsky.social · 26/10/2025
How can we design complex computational systems using synthetic biology? Check our paper, where we use space as a computational element, dramatically reducing communication requirements while allowing modularity, reusability, scalability & minimal cell engineering. journals.plos.org/ploscompbiol...
0368
Biplab Bose @biplabbose.bsky.social · 27/10/2025
Embryonic development is robust but fluctuations and variations are inevitable. Nature smartly sells those variations as features not bug. #systemsbiology #complexsystems
000
Biplab Bose @biplabbose.bsky.social · 26/10/2025
LLMs don’t live—yet they learn, adapt, and hallucinate. Strange! Is not it? Biologists need to rethink the key ideas of Life. #physicsoflife #mathbio #complexity #systemsbiology
020
Biplab Bose @biplabbose.bsky.social · 25/10/2025
Does a cell have a 'mind' - say a proto-mind or basal cognition? Although it was once a fringe idea, recent experimental and mathematical works are accumulating in its support. Here is an interesting recent work: www.biorxiv.org/content/10.1... #complexsystem #systemsbiology #sysbio
biorxiv.org
Engineering Basal Cognition: Minimal Genetic Circuits for Habituation, Sensitization, and Massed–Spaced Learning
Cognition is often associated with complex brains, yet many forms of learning—such as habituation, sensitization, and even spacing effects—have been observed in single cells and aneural organisms. The...
045
Reposted by Biplab Bose
NSF-Simons NITMB @nitmb.bsky.social · 24/10/2025
Data-driven modeling of complex biological systems Thank you for joining us at today's NITMB Seminar with Cody FitzGerald, James S. McDonnell Foundation Postdoctoral Fellow at Northwestern University! 📽️ Watch the seminar recording - www.youtube.com/watch?v=B2Y9...
101
Biplab Bose @biplabbose.bsky.social · 24/10/2025
If you have to design a 40 hr UG course on Computational Biology, what topics will you include in the syllabus? As I see across institutes, this course is heavily biased to protein structure, molecular dynamics simulations and drug design. #compbio #qbio
000
Biplab Bose @biplabbose.bsky.social · 23/10/2025
f(x::Int, y::Int) = x + y f(x::String, y::String) = x * y # In Julia, even functions believe in pluralism. #JuliaLang #FunInComputing #computingjokes #computing
020
Biplab Bose @biplabbose.bsky.social · 22/10/2025
Is it only me or for everyone- the feeds in the Bluesky Android app updates only when I use certain networks (ISPs) that too with lag. #bluesky #bsky
000
Biplab Bose @biplabbose.bsky.social · 22/10/2025
Switched to Julia 3 years ago. Still can’t uninstall MATLAB — held hostage by the interactive GUIs of dfield, pplane, and matcont. #JuliaLang #computing #dynamicalsystems
020