Towards Effective Generation of Synthetic Memory References Via Markovian Models

A TIME SERIES KNOWLEDGE MINING FRAMEWORK EXPLOITING THE SYNERGY BETWEEN SUBSEQUENCE CLUSTERING AND PREDICTIVE MARKOVIAN MODELS

FUZZY ECONOMIC REVIEW ◽

10.25102/fer.2009.01.03 ◽

2009 ◽

Vol 14 (01) ◽

Author(s):

Vasile Georgescu

Keyword(s):

Time Series ◽

Knowledge Mining ◽

Markovian Models

Download Full-text

Neural network aided approximation and parameter inference of non-Markovian models of gene expression

Nature Communications ◽

10.1038/s41467-021-22919-1 ◽

2021 ◽

Vol 12 (1) ◽

Author(s):

Qingchao Jiang ◽

Xiaoming Fu ◽

Shifu Yan ◽

Runlai Li ◽

Wenli Du ◽

...

Keyword(s):

Neural Network ◽

Stochastic Dynamics ◽

Stochastic Simulations ◽

Parameter Inference ◽

Biochemical Processes ◽

Markovian Models ◽

The Neural Network ◽

Large Numbers ◽

Small Set ◽

Noisy Measurements

AbstractNon-Markovian models of stochastic biochemical kinetics often incorporate explicit time delays to effectively model large numbers of intermediate biochemical processes. Analysis and simulation of these models, as well as the inference of their parameters from data, are fraught with difficulties because the dynamics depends on the system’s history. Here we use an artificial neural network to approximate the time-dependent distributions of non-Markovian models by the solutions of much simpler time-inhomogeneous Markovian models; the approximation does not increase the dimensionality of the model and simultaneously leads to inference of the kinetic parameters. The training of the neural network uses a relatively small set of noisy measurements generated by experimental data or stochastic simulations of the non-Markovian model. We show using a variety of models, where the delays stem from transcriptional processes and feedback control, that the Markovian models learnt by the neural network accurately reflect the stochastic dynamics across parameter space.

Download Full-text

Tailored generation of insulin producing cells from canine mesenchymal stem cells derived from bone marrow and adipose tissue

Scientific Reports ◽

10.1038/s41598-021-91774-3 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Watchareewan Rodprasert ◽

Sirirat Nantavisai ◽

Koranis Pathanachai ◽

Prasit Pavasant ◽

Thanaphum Osathanon ◽

...

Keyword(s):

Stem Cells ◽

Bone Marrow ◽

Mesenchymal Stem Cells ◽

Hanging Drop ◽

Differentiation Potential ◽

Insulin Producing Cells ◽

Edmonton Protocol ◽

Effective Generation ◽

Hanging Drop Culture ◽

Peptide Secretion

AbstractThe trend of regenerative therapy for diabetes in human and veterinary practices has conceptually been proven according to the Edmonton protocol and animal models. Establishing an alternative insulin-producing cell (IPC) resource for further clinical application is a challenging task. This study investigated IPC generation from two practical canine mesenchymal stem cells (cMSCs), canine bone marrow-derived MSCs (cBM-MSCs) and canine adipose-derived MSCs (cAD-MSCs). The results illustrated that cBM-MSCs and cAD-MSCs contain distinct pancreatic differentiation potential and require the tailor-made induction protocols. The effective generation of cBM-MSC-derived IPCs needs the integration of genetic and microenvironment manipulation using a hanging-drop culture of PDX1-transfected cBM-MSCs under a three-step pancreatic induction protocol. However, this protocol is resource- and time-consuming. Another study on cAD-MSC-derived IPC generation found that IPC colonies could be obtained by a low attachment culture under the three-step induction protocol. Further, Notch signaling inhibition during pancreatic endoderm/progenitor induction yielded IPC colonies through the trend of glucose-responsive C-peptide secretion. Thus, this study showed that IPCs could be obtained from cBM-MSCs and cAD-MSCs through different induction techniques. Also, further signaling manipulation studies should be conducted to maximize the protocol’s efficiency.

Download Full-text