Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Quantitative Biology. The day in quantitative biology: models, measurements and the data behind them.
Ines: I'm Ines.
Marcus: And I'm Marcus, here with Yuki.
Ines: We're looking at the paper: "Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events".
Marcus: Here's the gist of the paper. Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events investigates how cells can rapidly and accurately interpret the directionality of chemical cues at ultra-low chemoattractant concentrations by analyzing discrete receptor engagements governed by extreme first passage time statistics The aim of this paper is to explore the theoretical limits by which cells can establish the directionality of chemical cues at ultra-low chemoattractant concentrations A key cellular function is the ability to rapidly and accurately interpret the directionality of chemical or electric gradients to inform decisions on where to move, when to divide or whether to initiate a response to a threat The information regulating these decisions is encoded by the noisy binding activity of diffusing signals at membrane bound receptors Previous experiments utilized chemoattractant concentrations which saturated GPCRs However, as the precision of experimental measurement has expanded, it is now clear that GPCRs can be activated at remarkably low levels of chemoattractant concentrations At such low concentration levels, receptor engagement and consequently cellular signaling decision making are regulated by a small number of discrete binding events governed by extreme first passage time statistics Mathematical Framework for Sensing The analysis considers an idealized non-dimensional scenario where a circular cell is centered at the origin while a source of diffusing chemoattractant is placed at x0 = R
cos θ0,sin θ0: for R > 1 In this setting we assume at the initial time t = 0 that N molecules are released from the source and undergo unbiased random motion before eventually reaching a receptor on the cell surface Each of the N molecules released at the source eventually reaches the cell surface and binds a receptor, producing two observable quantities: the time of the binding event and its location on the cell surface Directional sensing is then an inverse problem: from the earliest binding events, what can the cell infer about the source?
Ines: The arrival times reflect the source distance R, while the arrival locations reflect the source direction θ0 Statistical Characterization of Arrival Times The most pertinent biological information on the source is its direction relative to the cell It was noticed in
20: that when recovering the location of a diffusive source from the time-dependent signal, the short time signal (0 < t ≪ 1) gave a strong estimate on the directionality of the source but yielded poor information on the distance The limiting distribution for P
Tk,N = t: is given in (11) and takes the form P
Tk,N = t: = 1/aN (k − 1)!
Marcus: exp
φ(t): , φ(t) = k(t − bN) / aN − exp t − bN / aN The variance in the impact location (38) is independent of the particle diffusivity D Estimators and Optimal Information Use The analysis shows that a cell can employ simple averaging over the first k arrivals to reduce the variance (error) in the accuracy Maximum likelihood estimation (MLE) can be used to form more accurate estimates of the source
34: The mean squared error E
(ρ res k) 2: is minimized at an optimal k = k∗ which satisfies a transcendental equation For N ≫ 1, the optimal number of receptor events is k∗ ≈ log N, independent of geometric factors which provide O(1) corrections The work derives and validates explicit asymptotic distributions for both the hitting times and hitting locations of diffusing signaling molecules at the cell surface The biological significance of this mathematical framework is that it provides a quantitative framework for comparing the directional information available to the cell under different hypothesized sensing mechanisms Our analysis demonstrates that directional information is strongly front-loaded in early arrivals, while later arrivals lose directional bias This provides a plausible mechanism that explains the experimentally observed rapid symmetry breaking and polarization responses on timescales much shorter than those required to establish steady-state concentration profiles Extreme Interarrival Times are Exponential The times T1,N − bN / aN,..., Tk,N − bN / aN →d (X1,..., Xk) as N → ∞ Hence, 2(X3 − X2) and (X2 − X1) are i.i.d.
Ines: exponential random variables with unit means The interarrival times convey much less information to the cell than the full arrival times, but exploiting the full arrival record requires nonlinear processing of the arrival times Recovery of Source Direction from Extreme Arrivals in 2D The resultant formed by averaging over a vector of each arrival, n res ≡ 1/k X k j=1 nj, nj = cos θj,sin θj be the equally weighted average of the first k impacts at the cell The error ρ res k in the resultant vector formed by k arrivals is described by the generalized chi-squared distribution
48: Pruning redundant reasoning tokens The single most informative sentence about the paper's subject and finding is that early binding events carry disproportionately more information about source directionality than later arrivals How the training signal is built The key mathematical result of this work is a derivation of the limiting distribution of the first k arrivals in this extreme statistics regime Specifically, in Sec.
Title and authors: Ines: Let's talk about what they actually did in this paper, "Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events."
Marcus: The title itself points to the core idea: using single cells to sense directionality when the chemical concentration is incredibly low. It’s not about saturating receptors anymore.
Ines: Exactly. They're moving away from models where everything is just a continuous, saturated signal and focusing on the discrete nature of binding events under extreme conditions.
Marcus: The paper sets up this scenario with N molecules diffusing from a source at distance R greater than one times the cell radius, and they are looking at the first k impacts.
Ines: They derive analytic expressions for the joint distribution of those receptor binding times and binding locations, which is pretty deep work for this area.
Marcus: And what they show is that while later arrivals lose their directional bias, those initial early events still give us a strong handle on the source directionality.
Ines: So it's like the cell gets a quick snapshot of the source direction before everything settles into a steady state.
Marcus: And they propose using an optimal number of receptor events, k star, which has asymptotic behavior around log N to minimize error in localization estimates.
Ines: That means we don't just take the first few impacts randomly; there's a specific statistical way to sample that gives you the best accuracy for your source location.
Marcus: It connects the discrete binding statistics directly to a practical strategy for how a cell should process that noisy input data.
The paper's summary: Ines: So, summarizing this paper, it’s about establishing the theoretical limits of directional sensing in these ultra-low concentration regimes.
Marcus: They take the non-dimensional setup—circular cell at origin, source at distance R greater than one—and they analyze what we can actually infer from those first k hits.
Ines: The big takeaway is that the information about the source direction is front-loaded into the earliest arrivals, and later events don't add much directional value.
Marcus: They show that arrival times give us a sense of the source distance R, while arrival locations give us information about the source direction theta zero.
Ines: And they show how you can use a weighted average of these first k impacts to get a result with reduced variance in your estimate of the source vector.
Marcus: They also detail how we can recover that direction from time differences, showing that the average of those angle differences has a specific distribution dependent on D and R.
Ines: It’s about providing a quantitative framework for comparing different sensing strategies, whether you use simple averaging or more complex maximum likelihood estimation.
Marcus: The paper validates that this mathematical description holds true in the biologically relevant pre-equilibrium regime, which is crucial because it's not the steady state we usually model.
The paper's improvements: Ines: Looking at what they suggest as an improvement, they move beyond just saying early events are better by deriving specific limiting distributions for the first k arrivals.
Marcus: Specifically, they find that the impact location of the k-th arrival has a limiting behavior theta k N that depends on N and R in a specific way.
Ines: The variance of this impact location, sigma squared over k N, is shown to be independent of the particle diffusivity D, which is interesting because diffusivity usually plays a big role in motion studies.
Marcus: That independence happens because there's a balance: the angular spread of the flux grows with diffusivity, but the arrival time scales inversely with diffusivity.
Ines: So, higher diffusion means molecules arrive sooner, but they sample the same range of angles as those arriving later, so D sets a timescale but not the geometry of those early impacts.
Marcus: They also show that exploiting the full set of arrival times allows for a far more accurate estimate of the source distance R, although that requires solving a nonlinear system.
Conclusion: Ines: So to wrap up on this paper, the main implication is that we have a solid mathematical way to compare different methods for inferring source directionality from noisy early signals.
Marcus: They’ve shown that the directional information is strongly front-loaded in those first few arrivals, which explains why we see rapid symmetry breaking and polarization responses on timescales much shorter than what it would take to build up a steady state.
Ines: Yuki, from a population genetics perspective, what does this mean for understanding how species might sense gradients in the wild?
Yuki: It supports the idea that even at very low concentrations, directional sensing is possible and rapid, which could explain how populations respond quickly to subtle environmental cues in their surroundings.
Marcus: And from a cohort perspective, it tells us that if we look at early binding events statistically, we can use an optimal number of impacts around log N to get the best localization estimate.
Ines: It’s a lot of detail, but the paper "Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events" gives us a powerful tool to model how cells rapidly interpret chemical gradients.
Department of Applied and Computational Mathematics and Statistics, University of Notre Dame · Department of Mathematics, University of Utah
q-bio.CB
Submitted: 2026-03-10
Updated: 2026-08-11
Journal ref: Journal of Mathematical Biology, Volume 93, No. 61 (2026)
DOI: 10.1007/s00285-026-02478-z
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 84/100
The gist: Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events investigates how cells can rapidly and accurately interpret the directionality of…
Key concepts
- Extreme First-Passage Time Statistics
- This refers to the statistical behavior of molecules diffusing randomly until they hit a receptor. At very low concentrations, only the very first binding events matter. These extreme statistics govern how quickly and where these initial interactions occur, which is key to rapid sensing.
- Arrival Times vs. Locations
- When a molecule binds, it gives two pieces of data: the time it arrived and where on the cell surface it bound. The arrival times are linked to the source distance (how far away), while the binding locations are linked to the source direction (which way it came from).
- Optimal Sampling Strategy
- Cells can be smarter than just looking at every impact. The analysis shows an optimal number of early events, k*, which is about log N. Using this optimal number minimizes errors when estimating the source location, balancing the need for more data against the risk of noise.
- Front-Loaded Directional Information
- The most crucial finding is that directional information is concentrated in the earliest binding events. Later impacts lose this directional bias. This suggests cells rapidly make decisions based only on the initial, high-information arrivals.
Terminology
Summary
Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events investigates how cells can rapidly and accurately interpret the directionality of chemical cues at ultra-low chemoattractant concentrations by analyzing discrete receptor engagements governed by extreme first passage time statistics
Introduction to the Problem
The aim of this paper is to explore the theoretical limits by which cells can establish the directionality of chemical cues at ultra-low chemoattractant concentrations A key cellular function is the ability to rapidly and accurately interpret the directionality of chemical or electric gradients to inform decisions on where to move, when to divide or whether to initiate a response to a threat The information regulating these decisions is encoded by the noisy binding activity of diffusing signals at membrane bound receptors Previous experiments utilized chemoattractant concentrations which saturated GPCRs However, as the precision of experimental measurement has expanded, it is now clear that GPCRs can be activated at remarkably low levels of chemoattractant concentrations At such low concentration levels, receptor engagement and consequently cellular signaling decision making are regulated by a small number of discrete binding events governed by extreme first passage time statistics
Mathematical Framework for Sensing
The analysis considers an idealized non-dimensional scenario where a circular cell is centered at the origin while a source of diffusing chemoattractant is placed at x0 = R[cos θ0,sin θ0] for R > 1 In this setting we assume at the initial time t = 0 that N molecules are released from the source and undergo unbiased random motion before eventually reaching a receptor on the cell surface Each of the N molecules released at the source eventually reaches the cell surface and binds a receptor, producing two observable quantities: the time of the binding event and its location on the cell surface Directional sensing is then an inverse problem: from the earliest binding events, what can the cell infer about the source? The arrival times reflect the source distance R, while the arrival locations reflect the source direction θ0
Statistical Characterization of Arrival Times
The most pertinent biological information on the source is its direction relative to the cell It was noticed in [20] that when recovering the location of a diffusive source from the time-dependent signal, the short time signal (0 < t ≪ 1) gave a strong estimate on the directionality of the source but yielded poor information on the distance The limiting distribution for P[Tk,N = t] is given in (11) and takes the form P[Tk,N = t] = 1/aN (k − 1)! exp[φ(t)], φ(t) = k(t − bN) / aN − exp t − bN / aN The variance in the impact location (38) is independent of the particle diffusivity D
Estimators and Optimal Information Use
The analysis shows that a cell can employ simple averaging over the first k arrivals to reduce the variance (error) in the accuracy Maximum likelihood estimation (MLE) can be used to form more accurate estimates of the source [34] The mean squared error E[(ρ res k) 2] is minimized at an optimal k = k∗ which satisfies a transcendental equation For N ≫ 1, the optimal number of receptor events is k∗ ≈ log N, independent of geometric factors which provide O(1) corrections
Conclusion and Biological Significance
The work derives and validates explicit asymptotic distributions for both the hitting times and hitting locations of diffusing signaling molecules at the cell surface The biological significance of this mathematical framework is that it provides a quantitative framework for comparing the directional information available to the cell under different hypothesized sensing mechanisms Our analysis demonstrates that directional information is strongly front-loaded in early arrivals, while later arrivals lose directional bias This provides a plausible mechanism that explains the experimentally observed rapid symmetry breaking and polarization responses on timescales much shorter than those required to establish steady-state concentration profiles
Extreme Interarrival Times are Exponential
The times T1,N − bN / aN,..., Tk,N − bN / aN →d (X1,..., Xk) as N → ∞ Hence, 2(X3 − X2) and (X2 − X1) are i. p. 3 Here W0(z) is the principal branch of the
LambertW function [33], which has behavior W0(z) = log z −log log z +· · · as z → ∞. Hence W0 ∼ 2 log N as N → ∞.Where the paper says it
Recovery of Source Direction from Extreme Arrivals in 2D
The resultant formed by averaging over a vector of each arrival, n res ≡ 1/k X k j=1 nj, nj = cos θj,sin θj be the equally weighted average of the first k impacts at the cell The error ρ res k in the resultant vector formed by k arrivals is described by the generalized chi-squared distribution [48]
Pruning redundant reasoning tokens
The single most informative sentence about the paper's subject and finding is that early binding events carry disproportionately more information about source directionality than later arrivals
How the training signal is built
The key mathematical result of this work is a derivation of the limiting distribution of the first k arrivals in this extreme statistics regime Specifically, in Sec. 4, we find that the impact location of the k th arrival to a circular cell has the limiting behavior θk,N ∼ N (θ0, σ2 k,N), σ2 k,N = 2D / R bN + aN log k
Where the gains come from
The variance in the impact location (38) is independent of the particle diffusivity D This reflects a cancellation of two competing effects: the angular spread of the flux conditioned on arrival at time t grows as 2Dtc/R while the characteristic arrival time of the k th molecule, tc = bN + aN log k, scales as 1/D A larger diffusivity therefore delivers the k th impact proportionally sooner by exactly the factor by which it broadens the angular distribution The result is that the k th arrival always samples the same range of angular space hence D sets the timescale of capture but not the geometry of the early-impact distribution
Discussion
The central mathematical contribution of this study is a derivation and validation of explicit asymptotic distributions for both the hitting times and hitting locations of diffusing signaling molecules at the cell surface By coupling extreme first-passage time statistics with short-time angular localization, we obtained a joint statistical description of early receptor binding events that is valid in the biologically relevant pre-equilibrium regime This enables a systematic comparison of inference strategies, including statistically optimal estimators and simpler biologically plausible estimators, on equal footng
References
[1] Copos, C., Sun, Y.-H., Zhu, K., Zhang, Y., Reid, B., Draper, B., Lin, F., Yue, H., Bernadskaya, Y., Zhao, M.: Galvanotactic directionality of cell groups depends on group size
[2] Henderson, N.T., Pablo, M., Ghose, D., Clark-Cotton, M.R., Zyla, T.R., Nolen, J., Elston, T.C., Lew, D.J.: Ratiometric gpcr signaling enables directional sensing in yeast
[3] Parent, C.A., Devreotes, P.N.: A cell’s sense of direction
[4] Sun, Y., Reid, B., Zhang, Y., Zhu, K., Ferreira, F., Estrada, A., Sun, Y., Draper, B.W., Yue, H., Copos, C.: Electric field–guided collective motility initiation of large epidermal cell groups
[5] Nwogbaga, I., Camley, B.A.: Cell shape and orientation control galvanotactic accuracy
[6] Kaiyrbekov, K., Camley, B.A.: Does nematic order allow groups of elongated cells to sense electric fields better? p. 1 1 Introduction
The aim of this paper is to explore the theoretical limits by which cells can establish
the directionality of chemical cues at ultra-low chemoattractant concentrations. A key
cellular function is the ability to rapidly and accurately interpret the directionality of
chemical or electric gradients to inform decisions on where to move, when to divide
or whether to initiate a response to a threat [1–6].Where the paper says it
[7] Kashyap, A., Wang, W., Camley, B.A.: Trade-offs in concentration sensing in dynamic environments. p. 0 edu;
Abstract
We investigate single-cell directional sensing from diffusing chemoattractant signals released by a localized source. We focus on the low-concentration regime in
which receptor activity is discrete and cellular decisions are made on timescales
far shorter than those required for steady-state concentration profiles or receptor
saturation to emerge.Where the paper says it
[8] Mugler, A., Levchenko, A., Nemenman, I.: Limits to the precision of gradient sensing with spatial communication and temporal integration. p. 0 Our results
demonstrate that cells possess sufficient information to rapidly and accurately
infer the directionality of sparse signals from a localized source through a small
number of discrete receptor engagements, well before a steady-state gradient has
been established. Keywords: Cell motility; Directional sensing; First passage times; Extreme statistics.Where the paper says it
[9] Camley, B.A.: Collective gradient sensing and chemotaxis: modeling and recent developments
[10] Berg, H.C.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Extreme directional sensing from sparse signals: The improved AI can rapidly infer source directionality from a small number of early receptor binding events, as shown by
an accurate directional estimate formed from an equally weighted average of the earliest impacts
being accurate even from a small number of impacts. -
Optimal sampling strategy for source localization: The AI can employ the optimal number of early binding events k star to minimize mean squared error, where
the critical number has asymptotic behavior k star ~ log N.
This allows the system to determine when more impacts no longer improve accuracy due to a bias-variance trade-off. -
Source distance and direction separation: The AI can distinguish between source distance and directionality by analyzing arrival times versus locations, noting that
the arrival times reflect the source distance, while the arrival locations reflect the source direction.
This provides a mechanism for separating these two pieces of information encoded in early binding data. -
Robust directional estimation from time differences: The system can use the average of angle differences between impacts to derive an estimate for source direction, as
the average of the differences then has the distribution N(0, 2D R k squared / (bN + aN log k))
whichimproves because fewer values are averaged.
-
High-precision source localization using full data: The AI can achieve vastly more accurate estimates of source distance by utilizing the entire set of arrival times, as
exploiting the full arrival times yields a far more accurate estimate of the source distance R,
although this requires solving a nonlinear system.
Abstract
We investigate single-cell directional sensing from diffusing chemoattractant signals released by a localized source. We focus on the low-concentration regime in which receptor activity is discrete and cellular decisions are made on timescales far shorter than those required for steady-state concentration profiles or receptor saturation to emerge. We derive analytic expressions for the joint distribution of receptor binding times and binding locations, conditional on the position of the source. We show that early binding events carry disproportionately more information about source directionality than later arrivals. Motivated by this observation, we propose and analyze several source-localization estimates that exploit early receptor binding statistics. For a directional estimate formed from an equally weighted average of the earliest impacts, we demonstrate that this estimate is accurate even from a small number of impacts, and derive an expression for the number of binding events that minimizes the mean squared error. Our results demonstrate that cells possess sufficient information to rapidly and accurately infer the directionality of sparse signals from a localized source through a small number of discrete receptor engagements, well before a steady-state gradient has been established.