Dataset Information

A biologist's guide to model selection and causal inference.

ABSTRACT: A goal of many research programmes in biology is to extract meaningful insights from large, complex datasets. Researchers in ecology, evolution and behavior (EEB) often grapple with long-term, observational datasets from which they construct models to test causal hypotheses about biological processes. Similarly, epidemiologists analyse large, complex observational datasets to understand the distribution and determinants of human health. A key difference in the analytical workflows for these two distinct areas of biology is the delineation of data analysis tasks and explicit use of causal directed acyclic graphs (DAGs), widely adopted by epidemiologists. Here, we review the most recent causal inference literature and describe an analytical workflow that has direct applications for EEB. We start this commentary by defining four distinct analytical tasks (description, prediction, association, causal inference). The remainder of the text is dedicated to causal inference, specifically focusing on the use of DAGs to inform the modelling strategy. Given the increasing interest in causal inference and misperceptions regarding this task, we seek to facilitate an exchange of ideas between disciplinary silos and provide an analytical framework that is particularly relevant for making causal inference from observational data.

SUBMITTER: Laubach ZM

PROVIDER: S-EPMC7893255 | biostudies-literature | 2021 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A biologist's guide to model selection and causal inference.

Laubach Zachary M ZM Murray Eleanor J EJ Hoke Kim L KL Safran Rebecca J RJ Perng Wei W

Proceedings. Biological sciences 20210127 1943

A goal of many research programmes in biology is to extract meaningful insights from large, complex datasets. Researchers in ecology, evolution and behavior (EEB) often grapple with long-term, observational datasets from which they construct models to test causal hypotheses about biological processes. Similarly, epidemiologists analyse large, complex observational datasets to understand the distribution and determinants of human health. A key difference in the analytical workflows for these two ...[more]

PMID: 33499782

Similar Datasets

Project description:Aurora kinases are essential for cell division and are frequently misregulated in human cancers. Based on their potential as cancer therapeutics, a plethora of small molecule Aurora kinase inhibitors have been developed, with a subset having been adopted as tools in cell biology. Here, we fill a gap in the characterization of Aurora kinase inhibitors by using biochemical and cell-based assays to systematically profile a panel of 10 commercially available compounds with reported selectivity for Aurora A (MLN8054, MLN8237, MK-5108, MK-8745, Genentech Aurora Inhibitor 1), Aurora B (Hesperadin, ZM447439, AZD1152-HQPA, GSK1070916), or Aurora A/B (VX-680). We quantify the in vitro effect of each inhibitor on the activity of Aurora A alone, as well as Aurora A and Aurora B bound to fragments of their activators, TPX2 and INCENP, respectively. We also report kinome profiling results for a subset of these compounds to highlight potential off-target effects. In a cellular context, we demonstrate that immunofluorescence-based detection of LATS2 and histone H3 phospho-epitopes provides a facile and reliable means to assess potency and specificity of Aurora A versus Aurora B inhibition, and that G2 duration measured in a live imaging assay is a specific readout of Aurora A activity. Our analysis also highlights variation between HeLa, U2OS, and hTERT-RPE1 cells that impacts selective Aurora A inhibition. For Aurora B, all four tested compounds exhibit excellent selectivity and do not significantly inhibit Aurora A at effective doses. For Aurora A, MK-5108 and MK-8745 are significantly more selective than the commonly used inhibitors MLN8054 and MLN8237. A crystal structure of an Aurora A/MK-5108 complex that we determined suggests the chemical basis for this higher specificity. Taken together, our quantitative biochemical and cell-based analyses indicate that AZD1152-HQPA and MK-8745 are the best current tools for selectively inhibiting Aurora B and Aurora A, respectively. However, MK-8745 is not nearly as ideal as AZD1152-HQPA in that it requires high concentrations to achieve full inhibition in a cellular context, indicating a need for more potent Aurora A-selective inhibitors. We conclude with a set of "good practice" guidelines for the use of Aurora inhibitors in cell biology experiments.

Dataset Information

A biologist's guide to model selection and causal inference.

Publications

A biologist's guide to model selection and causal inference.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure