© 2026 Optics and Photonics at Nottingham
Publications
in Top 10% Journal Percentiles
43%
Global
Partnerships
56.3%
Corporate
Collaboration
8.8%
Annual research
funding
£5m+
Data for 2020-2025 from SciVal
Selected Publications
2024
3.
Burrows, Liam; Sculthorpe, Declan; Zhang, Hongrun; Rehman, Obaid; Mukherjee, Abhik; Chen, Ke
In: J. Pathol. Inform., vol. 15, no. 100351, pp. 100351, 2024.
Abstract | Links | Altmetric | Tags: Digital multiplex, Digital pathology, Machine learning, Mathematical modelling, Stromal stain, Tissue microarrays
@article{Burrows2024-xb,
title = {Mathematical modelling and deep learning algorithms to automate assessment of single and digitally multiplexed immunohistochemical stains in tumoural stroma},
author = {Liam Burrows and Declan Sculthorpe and Hongrun Zhang and Obaid Rehman and Abhik Mukherjee and Ke Chen},
doi = {10.1016/j.jpi.2023.100351},
year = {2024},
date = {2024-12-01},
urldate = {2024-12-01},
journal = {J. Pathol. Inform.},
volume = {15},
number = {100351},
pages = {100351},
publisher = {Elsevier BV},
abstract = {Whilst automated analysis of immunostains in pathology research
has focused predominantly on the epithelial compartment,
automated analysis of stains in the stromal compartment is
challenging and therefore requires time-consuming pathological
input and guidance to adjust to tissue morphometry as perceived
by pathologists. This study aimed to develop a robust method to
automate stromal stain analyses using 2 of the commonest stromal
stains (SMA and desmin) employed in clinical pathology practice
as examples. An effective computational method capable of
automatically assessing and quantifying tumour-associated
stromal stains was developed and applied on cores of colorectal
cancer tissue microarrays. The methodology combines both
mathematical models and deep learning techniques with the former
requiring no training data and the latter as many inputs as
possible. The novel mathematical model was used to produce a
digital double marker overlay allowing for fast automated
digital multiplex analysis of stromal stains. The results show
that deep learning methodologies in combination with
mathematical modelling allow for an accurate means of
quantifying stromal stains whilst also opening up new
possibilities of digital multiplex analyses.},
keywords = {Digital multiplex, Digital pathology, Machine learning, Mathematical modelling, Stromal stain, Tissue microarrays},
pubstate = {published},
tppubtype = {article}
}
Whilst automated analysis of immunostains in pathology research
has focused predominantly on the epithelial compartment,
automated analysis of stains in the stromal compartment is
challenging and therefore requires time-consuming pathological
input and guidance to adjust to tissue morphometry as perceived
by pathologists. This study aimed to develop a robust method to
automate stromal stain analyses using 2 of the commonest stromal
stains (SMA and desmin) employed in clinical pathology practice
as examples. An effective computational method capable of
automatically assessing and quantifying tumour-associated
stromal stains was developed and applied on cores of colorectal
cancer tissue microarrays. The methodology combines both
mathematical models and deep learning techniques with the former
requiring no training data and the latter as many inputs as
possible. The novel mathematical model was used to produce a
digital double marker overlay allowing for fast automated
digital multiplex analysis of stromal stains. The results show
that deep learning methodologies in combination with
mathematical modelling allow for an accurate means of
quantifying stromal stains whilst also opening up new
possibilities of digital multiplex analyses.
has focused predominantly on the epithelial compartment,
automated analysis of stains in the stromal compartment is
challenging and therefore requires time-consuming pathological
input and guidance to adjust to tissue morphometry as perceived
by pathologists. This study aimed to develop a robust method to
automate stromal stain analyses using 2 of the commonest stromal
stains (SMA and desmin) employed in clinical pathology practice
as examples. An effective computational method capable of
automatically assessing and quantifying tumour-associated
stromal stains was developed and applied on cores of colorectal
cancer tissue microarrays. The methodology combines both
mathematical models and deep learning techniques with the former
requiring no training data and the latter as many inputs as
possible. The novel mathematical model was used to produce a
digital double marker overlay allowing for fast automated
digital multiplex analysis of stromal stains. The results show
that deep learning methodologies in combination with
mathematical modelling allow for an accurate means of
quantifying stromal stains whilst also opening up new
possibilities of digital multiplex analyses.
2023
2.
Pérez-Cota, Fernando; Martínez-Arellano, Giovanna; 3rd La Cavera, Salvatore; Hardiman, William; Thornton, Luke; Fuentes-Domínguez, Rafael; Smith, Richard J; McIntyre, Alan; Clark, Matt
Classification of cancer cells at the sub-cellular level by phonon microscopy using deep learning Journal Article
In: Sci. Rep., vol. 13, no. 1, pp. 16228, 2023.
Abstract | Tags: Deep learning, Machine learning, microscopy, Phonon microscopy
@article{Perez-Cota2023-nu,
title = {Classification of cancer cells at the sub-cellular level by phonon microscopy using deep learning},
author = {Fernando P\'{e}rez-Cota and Giovanna Mart\'{i}nez-Arellano and Salvatore 3rd La Cavera and William Hardiman and Luke Thornton and Rafael Fuentes-Dom\'{i}nguez and Richard J Smith and Alan McIntyre and Matt Clark},
year = {2023},
date = {2023-09-01},
urldate = {2023-09-01},
journal = {Sci. Rep.},
volume = {13},
number = {1},
pages = {16228},
publisher = {Springer Science and Business Media LLC},
abstract = {There is a consensus about the strong correlation between the
elasticity of cells and tissue and their normal, dysplastic, and
cancerous states. However, developments in cell mechanics have
not seen significant progress in clinical applications. In this
work, we explore the possibility of using phonon acoustics for
this purpose. We used phonon microscopy to obtain a measure of
the elastic properties between cancerous and normal breast
cells. Utilising the raw time-resolved phonon-derived data (300
k individual inputs), we employed a deep learning technique to
differentiate between MDA-MB-231 and MCF10a cell lines. We
achieved a 93% accuracy using a single phonon measurement in a
volume of approximately 2.5 μm3. We also investigated means
for classification based on a physical model that suggest the
presence of unidentified mechanical markers. We have
successfully created a compact sensor design as a proof of
principle, demonstrating its compatibility for use with needles
and endoscopes, opening up exciting possibilities for future
applications.},
keywords = {Deep learning, Machine learning, microscopy, Phonon microscopy},
pubstate = {published},
tppubtype = {article}
}
There is a consensus about the strong correlation between the
elasticity of cells and tissue and their normal, dysplastic, and
cancerous states. However, developments in cell mechanics have
not seen significant progress in clinical applications. In this
work, we explore the possibility of using phonon acoustics for
this purpose. We used phonon microscopy to obtain a measure of
the elastic properties between cancerous and normal breast
cells. Utilising the raw time-resolved phonon-derived data (300
k individual inputs), we employed a deep learning technique to
differentiate between MDA-MB-231 and MCF10a cell lines. We
achieved a 93% accuracy using a single phonon measurement in a
volume of approximately 2.5 μm3. We also investigated means
for classification based on a physical model that suggest the
presence of unidentified mechanical markers. We have
successfully created a compact sensor design as a proof of
principle, demonstrating its compatibility for use with needles
and endoscopes, opening up exciting possibilities for future
applications.
elasticity of cells and tissue and their normal, dysplastic, and
cancerous states. However, developments in cell mechanics have
not seen significant progress in clinical applications. In this
work, we explore the possibility of using phonon acoustics for
this purpose. We used phonon microscopy to obtain a measure of
the elastic properties between cancerous and normal breast
cells. Utilising the raw time-resolved phonon-derived data (300
k individual inputs), we employed a deep learning technique to
differentiate between MDA-MB-231 and MCF10a cell lines. We
achieved a 93% accuracy using a single phonon measurement in a
volume of approximately 2.5 μm3. We also investigated means
for classification based on a physical model that suggest the
presence of unidentified mechanical markers. We have
successfully created a compact sensor design as a proof of
principle, demonstrating its compatibility for use with needles
and endoscopes, opening up exciting possibilities for future
applications.
2022
1.
Jiménez-Rosés, Mireia; Morgan, Bradley Angus; Sigstad, Maria Jimenez; Tran, Thuy Duong Zoe; Srivastava, Rohini; Bunsuz, Asuman; Borrega-Román, Leire; Hompluem, Pattarin; Cullum, Sean A; Harwood, Clare R; Koers, Eline J; Sykes, David A; Styles, Iain B; Veprintsev, Dmitry B
Combined docking and machine learning identify key molecular determinants of ligand pharmacological activity on β2 adrenoceptor Journal Article
In: Pharmacol. Res. Perspect., vol. 10, no. 5, pp. e00994, 2022.
Abstract | Tags: adrenoceptor, docking, drug discovery, GPCRs, Machine learning, structure-activity relationship
@article{Jimenez-Roses2022-ss,
title = {Combined docking and machine learning identify key molecular
determinants of ligand pharmacological activity on β2
adrenoceptor},
author = {Mireia Jim\'{e}nez-Ros\'{e}s and Bradley Angus Morgan and Maria Jimenez Sigstad and Thuy Duong Zoe Tran and Rohini Srivastava and Asuman Bunsuz and Leire Borrega-Rom\'{a}n and Pattarin Hompluem and Sean A Cullum and Clare R Harwood and Eline J Koers and David A Sykes and Iain B Styles and Dmitry B Veprintsev},
year = {2022},
date = {2022-10-01},
journal = {Pharmacol. Res. Perspect.},
volume = {10},
number = {5},
pages = {e00994},
publisher = {Wiley},
abstract = {G protein-coupled receptors (GPCRs) are valuable therapeutic
targets for many diseases. A central question of GPCR drug
discovery is to understand what determines the agonism or
antagonism of ligands that bind them. Ligands exert their action
via the interactions in the ligand binding pocket. We
hypothesized that there is a common set of receptor interactions
made by ligands of diverse structures that mediate their action
and that among a large dataset of different ligands, the
functionally important interactions will be over-represented. We
computationally docked ~2700 known β2AR ligands to
multiple β2AR structures, generating ca 75 000 docking
poses and predicted all atomic interactions between the receptor
and the ligand. We used machine learning (ML) techniques to
identify specific interactions that correlate with the agonist
or antagonist activity of these ligands. We demonstrate with the
application of ML methods that it is possible to identify the
key interactions associated with agonism or antagonism of
ligands. The most representative interactions for agonist
ligands involve K972.68$times$67 , F194ECL2 ,
S2035.42$times$43 , S2045.43$times$44 , S2075.46$times$641 ,
H2966.58$times$58 , and K3057.32$times$31 . Meanwhile, the
antagonist ligands made interactions with W2866.48$times$48 and
Y3167.43$times$42 , both residues considered to be important in
GPCR activation. The interpretation of ML analysis in human
understandable form allowed us to construct an exquisitely
detailed structure-activity relationship that identifies small
changes to the ligands that invert their pharmacological
activity and thus helps to guide the drug discovery process.
This approach can be readily applied to any drug target.},
keywords = {adrenoceptor, docking, drug discovery, GPCRs, Machine learning, structure-activity relationship},
pubstate = {published},
tppubtype = {article}
}
G protein-coupled receptors (GPCRs) are valuable therapeutic
targets for many diseases. A central question of GPCR drug
discovery is to understand what determines the agonism or
antagonism of ligands that bind them. Ligands exert their action
via the interactions in the ligand binding pocket. We
hypothesized that there is a common set of receptor interactions
made by ligands of diverse structures that mediate their action
and that among a large dataset of different ligands, the
functionally important interactions will be over-represented. We
computationally docked ~2700 known β2AR ligands to
multiple β2AR structures, generating ca 75 000 docking
poses and predicted all atomic interactions between the receptor
and the ligand. We used machine learning (ML) techniques to
identify specific interactions that correlate with the agonist
or antagonist activity of these ligands. We demonstrate with the
application of ML methods that it is possible to identify the
key interactions associated with agonism or antagonism of
ligands. The most representative interactions for agonist
ligands involve K972.68$times$67 , F194ECL2 ,
S2035.42$times$43 , S2045.43$times$44 , S2075.46$times$641 ,
H2966.58$times$58 , and K3057.32$times$31 . Meanwhile, the
antagonist ligands made interactions with W2866.48$times$48 and
Y3167.43$times$42 , both residues considered to be important in
GPCR activation. The interpretation of ML analysis in human
understandable form allowed us to construct an exquisitely
detailed structure-activity relationship that identifies small
changes to the ligands that invert their pharmacological
activity and thus helps to guide the drug discovery process.
This approach can be readily applied to any drug target.
targets for many diseases. A central question of GPCR drug
discovery is to understand what determines the agonism or
antagonism of ligands that bind them. Ligands exert their action
via the interactions in the ligand binding pocket. We
hypothesized that there is a common set of receptor interactions
made by ligands of diverse structures that mediate their action
and that among a large dataset of different ligands, the
functionally important interactions will be over-represented. We
computationally docked ~2700 known β2AR ligands to
multiple β2AR structures, generating ca 75 000 docking
poses and predicted all atomic interactions between the receptor
and the ligand. We used machine learning (ML) techniques to
identify specific interactions that correlate with the agonist
or antagonist activity of these ligands. We demonstrate with the
application of ML methods that it is possible to identify the
key interactions associated with agonism or antagonism of
ligands. The most representative interactions for agonist
ligands involve K972.68$times$67 , F194ECL2 ,
S2035.42$times$43 , S2045.43$times$44 , S2075.46$times$641 ,
H2966.58$times$58 , and K3057.32$times$31 . Meanwhile, the
antagonist ligands made interactions with W2866.48$times$48 and
Y3167.43$times$42 , both residues considered to be important in
GPCR activation. The interpretation of ML analysis in human
understandable form allowed us to construct an exquisitely
detailed structure-activity relationship that identifies small
changes to the ligands that invert their pharmacological
activity and thus helps to guide the drug discovery process.
This approach can be readily applied to any drug target.
A part of the University of Nottingham
© 2026 Optics and Photonics at Nottingham. Created for free using WordPress and Kubio