Biology subjects

Filby, A.

Publications and source records attributed to Filby, A..

23 records · Page 2Linked to original sources

bioRxiv2019

Classification of human white blood cells using machine learning for stain-free imaging flow cytometry

Imaging flow cytometry (IFC) produces up to 12 different information-rich images of single cells at a throughput of 5000 cells per second. Yet often, cell populations are still studied using manual gating, a technique that has several drawbacks. Firstly, it is hard to reproduce. Secondly, it is subjective and biased. And thirdly, it is time-consuming for large experiments. Therefore, it would be advantageous to replace manual gating with an automated process, which could be based on stain-free measurements originating from the brightfield and darkfield image channels. To realise this potential, advanced data analysis methods are required, in particular, machine learning. Previous works have successfully tested this approach on cell cycle phase classification with both a classical machine learning approach based on manually engineered features, and a deep learning approach. In this work, we compare both approaches extensively on the complex problem of white blood cell classification. Four human whole blood samples were assayed on an ImageStream-X MK II imaging flow cytometer. Two samples were stained for the identification of 8 white blood cell types, while two other sample sets were stained for the identification of resting and active eosinophils. For both datasets, four machine learning classifiers were evaluated on stain-free imagery using stratified 5-fold cross-validation. On the white blood cell dataset the best obtained results were 0.776 and 0.697 balanced accuracy for classical machine learning and deep learning, respectively. On the eosinophil dataset this was 0.866 and 0.867 balanced accuracy. From the experiments we conclude that classifying distinct cell types based on only stain-free images is possible with these techniques. However, both approaches did not always succeed in making reliable cell subtype classifications. Also, depending on the cell type, we find that even though the deep learning approach requires less expert input, it performs on par with a classical approach.

bioinformatics↗

bioRxiv2019

Decoding the development of the blood and immune systems during human fetal liver haematopoiesis

Definitive haematopoiesis in the fetal liver supports self-renewal and differentiation of haematopoietic stem cells/multipotent progenitors (HSC/MPPs), yet remains poorly defined in humans. Using single cell transcriptome profiling of ~133,000 fetal liver and ~65,000 fetal skin and kidney cells, we identify the repertoire of blood and immune cells in first and early second trimesters of development. From this data, we infer differentiation trajectories from HSC/MPPs, and evaluate the impact of tissue microenvironment on blood and immune cell development. We predict coupling of mast cell differentiation with erythro-megakaryopoiesis and identify physiological erythropoiesis in fetal skin. We demonstrate a shift in fetal liver haematopoietic composition during gestation away from being erythroid-predominant, accompanied by a parallel change in HSC/MPP differentiation potential, which we functionally validate. Our integrated map of fetal liver haematopoiesis provides a blueprint for the study of paediatric blood and immune disorders, and a valuable reference for understanding and harnessing the therapeutic potential of HSC/MPPs.

developmental biology↗

bioRxiv2018

Reconstructing the human first trimester fetal-maternal interface using single cell transcriptomics

During the early weeks of human pregnancy, the fetal placenta implants into the uterine mucosa (decidua) where placental trophoblast cells intermingle and communicate with maternal cells. Here, we profile transcriptomes of [~]50,000 single cells from this unique microenvironment, sampling matched first trimester maternal blood and decidua, and fetal cells from the placenta itself. We define the cellular composition of human decidua, revealing five distinct subsets of decidual fibroblasts with differing growth factors and hormone production profiles, and show that fibroblast states define two distinct decidual layers. Among decidual NK cells, we resolve three subsets, each with a different immunomodulatory and chemokine profile. We develop a repository of ligand-receptor pairs (www.CellPhoneDB.org) and a statistical tool to predict the probability of cell-cell interactions via these pairs, highlighting specific interactions between decidual NK cells and invading fetal extravillous trophoblast cells, maternal immune and stromal cells. Our single cell atlas of the maternal-fetal interface reveals the cellular organization and interactions critical for placentation and reproductive success.

developmental biology↗

bioRxiv2017

A Universal Metric For Evaluating, Optimising And Benchmarking The Performance Of A Research Technology Platform (RTP)

Research Technology Platforms (RTPs) exist to facilitate the application and utilisation of specific analytical technologies to the highest possible standard thus delivering reputable data across a broad spectrum of research themes. Specifically, RTPs centralise expertise in a given technology and provide an unparalleled level of continuity and practical knowledge retention that simply cannot be achieved by more organic, ad hoc means of support. As small non profit businesses often tasked with recovering all or a percentage of their running costs, RTPs are under significant pressure to keep pace with rapidly advancing technology and new methodologies against a back drop of dwindling funding for scientific research. At present there are a number of non-trivial issues that make assessing the operational performance of a RTP difficult to determine on a standalone basis let alone attempting to benchmark against other RTPs within the same or different technology fields. Firstly, depending on the technological speciality the RTP may work to one of essentially three operational models. RTPs such as Bio-Imaging or Cytometry provide access to well-maintained analytical systems that can be utilised by trained individuals for a timed access charge. In some cases there will be a requirement for assisted operation of certain instruments by core staff (e.g. cell sorters). Genomics and Proteomics RTPs tend to function on a project basis whereby users will not access the technology themselves rather pay for a full analytical service often with a milestone-based approach for tracking progress. Other RTPs work to a hybrid approach were technical staff provide certain elements of sample preparation for specific projects prior to analysis on core supported, user accessible instrumentation. Secondly the specific operational costs that each RTP is tasked to recover varies significantly on a local, national and international level due to institutional subsidies. These operational costs can include staff salaries, instrument maintenance, associated running consumables, and in some cases instrument depreciation but there is standardised rule as to what each RTP is tasked to recover and to what percentage.\n\nHere we present a generalised mathematical approach to describe the customisable metrics of any given RTP serviceThe general strategy how to increase performance within the framework of this approach has been identified through breaking down these customisable metrics into components and maximising them according to specific requirements. These strategies could be potentially adopted for different operational or local procedures, integrating the specifics related to the institutional or national policies. The approach laid down here should be considered as a trigger for opening a discussion around how to address optimising RTP performance and allow for benchmarking across the full breadth of RTPs.

scientific communication and education↗

bioRxiv2016

Deep Learning for Imaging Flow Cytometry: Cell Cycle Analysis of Jurkat Cells

We show that deep convolutional neural networks combined with non-linear dimension reduction enable reconstructing biological processes based on raw image data. We demonstrate this by recon-structing the cell cycle of Jurkat cells and disease progression in diabetic retinopathy. In further analysis of Jurkat cells, we detect and separate a subpopulation of dead cells in an unsupervised manner and, in classifying discrete cell cycle stages, we reach a 6-fold reduction in error rate compared to a recent approach based on boosting on image features. In contrast to previous methods, deep learning based predictions are fast enough for on-the-fly analysis in an imaging flow cytometer.

cell biology↗