Few-shot deep learning for AFM force curve characterization of single-molecule interactions

Waite, Joshua R.; Tan, Sin Yong; Saha, Homagni; Sarkar, Soumik; Sarkar, Anwesha

doi:10.1016/j.patter.2022.100672

Cited by 9 publications

(6 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With sophisticated computational approaches slowly gaining traction in AFM studies, errors in data processing may become less prevalent and reduce the time spent analyzing data. Notably, finite and inverse finite element models [ 47 , 65 , 66 ] and machine learning algorithms [ 9 , 67 , 68 , 69 ] have been developed and implemented to analyze AFM nanoindentation data more accurately. Minelli et al, for example, were able to discriminate healthy tissues from cancer tissues using a fully automated neural network analysis that evaluates force–distance curves [ 69 ].…”

Section: Discussionmentioning

confidence: 99%

“…AFM is also superior in its ability to provide high-resolution nanotopographical images and requires relatively simple sample preparation, although special considerations must be made for accurate mechanical analysis [ 5 , 6 , 7 , 8 ]. Compared to most other techniques, however, AFM is low-throughput, time-consuming, and technically challenging [ 6 , 7 , 9 ]. Nonetheless, AFM has led to pioneering work that has advanced our understanding of cell and tissue mechanics, particularly in the context of cancer.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Atomic Force Microscopy Methods to Measure Tumor Mechanical Properties

Najera

Rosenberger

Datta

2023

Cancers

View full text Add to dashboard Cite

Atomic force microscopy (AFM) is a popular tool for evaluating the mechanical properties of biological materials (cells and tissues) at high resolution. This technique has become particularly attractive to cancer researchers seeking to bridge the gap between mechanobiology and cancer initiation, progression, and treatment resistance. The majority of AFM studies thus far have been extensively focused on the nanomechanical characterization of cells. However, these approaches fail to capture the complex and heterogeneous nature of a tumor and its host organ. Over the past decade, efforts have been made to characterize the mechanical properties of tumors and tumor-bearing tissues using AFM. This has led to novel insights regarding cancer mechanopathology at the tissue scale. In this Review, we first explain the principles of AFM nanoindentation for the general study of tissue mechanics. We next discuss key considerations when using this technique and preparing tissue samples for analysis. We then examine AFM application in characterizing the mechanical properties of cancer tissues. Finally, we provide an outlook on AFM in the field of cancer mechanobiology and its application in the clinic.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Atomic Force Microscopy Methods to Measure Tumor Mechanical Properties

Najera

Rosenberger

Datta

2023

Cancers

View full text Add to dashboard Cite

show abstract

“…In addition, there is also the chance of the model overfitting to noise in the data. In common ML applications, the number of instances in a dataset usually goes from a few thousand to hundreds of thousands, and it is no different when applied to AFM frameworks [31,46]. Thus, to ensure a robust initial model without excessively compromising computational costs, an initial dataset consisting of 40,000 curves was created.…”

Section: Synthetic Nanoindentation Curvesmentioning

confidence: 99%

Deep Learning Regressors of Surface Properties from Atomic Force Microscopy Nanoindentations

Pacheco,

Ferreira,

Parente

2024

Applied Sciences

View full text Add to dashboard Cite

Atomic force microscopy (AFM) is a powerful technique to study the nanomechanical properties of a wide range of materials at the piconewton level. AFM force–indentation curves can be fitted with appropriate contact models, enabling the determination of material properties for a given sample. However, the analysis of large datasets comprising thousands of curves using conventional methods presents a time-intensive challenge. As a result, there is an increasing interest in exploring alternative methodologies, such as integrating machine learning (ML) models to streamline and improve the efficiency of this process. In this work, two data-driven regressors were tuned to predict the Young’s modulus and adhesion energy from force–indentation curves of soft samples (Young’s modulus up to 10 kPa). Both models were trained exclusively on synthetic data derived from the contact theories developed by Hertz as well as Johnson, Kendall and Roberts (JKR). The PyTorch library was employed to build and train the models; then, the key hyperparameters were refined by implementing the optimization framework Optuna. The first model was successfully tested with synthetic and experimental curves from AFM nanoindentations, and the second presented promising results on the synthetic data. Our work suggests that experimental data may not be essential for training data-driven models to predict surface properties from AFM nanoindentations. By delivering accurate predictions in a computationally efficient way, our regressors validate the potential of a deep learning approach in exploring AFM nanoindentations and motivate further development of similar strategies to overcome current limitations in AFM postprocessing.

show abstract

“…[20][21][22][23][24] However, its use for processing SMFS data curves is much more limited. A recent study 25 trained a 1D Convolutional Neural Network (CNN) 26 using a triplet loss function 27 to generate an embedding space where curves containing single, double and multiple or no rupture events were separated. The authors demonstrated 65-70% accuracy in classifying a subset of data traces that were reliably and consistently classified by experts.…”

Section: Introductionmentioning

confidence: 99%

Iterative Machine Learning for Classification and Discovery of Single-molecule Unfolding Trajectories from Force Spectroscopy Data

Doffini

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

We report the application of machine learning techniques to accelerate classification and analysis of protein unfolding trajectories from force spectroscopy data. Using kernel methods, logistic regression and triplet loss, we developed a workflow called Forced Unfolding and Supervised Iterative Online (FUSION) where a user classifies a small number of repeatable unfolding patterns encoded as image data, and a machine is tasked with identifying similar images to classify the remaining data. We tested the workflow using two case studies on a multi-domain XMod-Dockerin/Cohesin complex, validating the approach first using synthetic data generated with a Monte Carlo algorithm, and then deploying the method on experimental atomic force spectroscopy data. FUSION efficiently separated traces that passed quality filters from unusable ones, classified curves with high accuracy, and identified unfolding pathways undetected by the user. This study demonstrates the potential of machine learning to accelerate data analysis, and generate new insights in protein biophysics.

show abstract

Few-shot deep learning for AFM force curve characterization of single-molecule interactions

Cited by 9 publications

References 48 publications

Atomic Force Microscopy Methods to Measure Tumor Mechanical Properties

Atomic Force Microscopy Methods to Measure Tumor Mechanical Properties

Deep Learning Regressors of Surface Properties from Atomic Force Microscopy Nanoindentations

Iterative Machine Learning for Classification and Discovery of Single-molecule Unfolding Trajectories from Force Spectroscopy Data

Contact Info

Product

Resources

About