A quantitative approach for understanding small-scale human mesenchymal stem cell culture - implications for large-scale bioprocess development

Human mesenchymal stem cell (hMSC) therapies have the potential to revolutionise the healthcare industry and replicate the success of the therapeutic protein industry; however, for this to be achieved there is a need to apply key bioprocessing engineering principles and adopt a quantitative approach for large-scale reproducible hMSC bioprocess development. Here we provide a quantitative analysis of the changes in concentration of glucose, lactate and ammonium with time during hMSC monolayer culture over 4 passages, under 100% and 20% dissolved oxgen (dO2), where either a 100%, 50% or 0% growth medium exchange was performed after 72h in culture. Yield coefficients, specific growth rates (h-1) and doubling times (h) were calculated for all cases. The 100% dO2 flasks outperformed the 20% dO2 flasks with respect to cumulative cell number, with the latter consuming more glucose and producing more lactate and ammonium. Furthermore, the 100% and 50% medium exchange conditions resulted in similar cumulative cell numbers, whilst the 0% conditions were significantly lower. Cell immunophenotype and multipotency were not affected by the experimental culture conditions. This study demonstrates the importance of determining optimal culture conditions for hMSC expansion and highlights a potential cost savings from only making a 50% medium exchange, which may prove significant for large-scale bioprocessing. © 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

(Tables 1, 3) Details of Agassiz trawl stations and numbers of specimens per macro- and megazoobenthic taxon collected during POLARSTERN cruise ANT-XXII/3 (ANDEEP III)

The assemblages inhabiting the continental shelf around Antarctica are known to be very patchy, in large part due to deep iceberg impacts. The present study shows that richness and abundance of much deeper benthos, at slope and abyssal depths, also vary greatly in the Southern and South Atlantic oceans. On the ANDEEP III expedition, we deployed 16 Agassiz trawls to sample the zoobenthos at depths from 1055 to 4930 m across the northern Weddell Sea and two South Atlantic basins. A total of 5933 specimens, belonging to 44 higher taxonomic groups, were collected. Overall the most frequent taxa were Ophiuroidea, Bivalvia, Polychaeta and Asteroidea, and the most abundant taxa were Malacostraca, Polychaeta and Bivalvia. Species richness per station varied from 6 to 148. The taxonomic composition of assemblages, based on relative taxon richness, varied considerably between sites but showed no relation to depth. The former three most abundant taxa accounted for 10-30% each of all taxa present. Standardised abundances based on trawl catches varied between 1 and 252 individuals per 1000 m2. Abundance significantly decreased with increasing depth, and assemblages showed high patchiness in their distribution. Cluster analysis based on relative abundance showed changes of community structure that were not linked to depth, area, sediment grain size or temperature. Generally abundances of zoobenthos in the abyssal Weddell Sea are lower than shelf abundances by several orders of magnitude.

Bayesian Inference in Large-scale Problems

Many modern applications fall into the category of "large-scale" statistical problems, in which both the number of observations n and the number of features or parameters p may be large. Many existing methods focus on point estimation, despite the continued relevance of uncertainty quantification in the sciences, where the number of parameters to estimate often exceeds the sample size, despite huge increases in the value of n typically seen in many fields. Thus, the tendency in some areas of industry to dispense with traditional statistical analysis on the basis that "n=all" is of little relevance outside of certain narrow applications. The main result of the Big Data revolution in most fields has instead been to make computation much harder without reducing the importance of uncertainty quantification. Bayesian methods excel at uncertainty quantification, but often scale poorly relative to alternatives. This conflict between the statistical advantages of Bayesian procedures and their substantial computational disadvantages is perhaps the greatest challenge facing modern Bayesian statistics, and is the primary motivation for the work presented here.

Two general strategies for scaling Bayesian inference are considered. The first is the development of methods that lend themselves to faster computation, and the second is design and characterization of computational algorithms that scale better in n or p. In the first instance, the focus is on joint inference outside of the standard problem of multivariate continuous data that has been a major focus of previous theoretical work in this area. In the second area, we pursue strategies for improving the speed of Markov chain Monte Carlo algorithms, and characterizing their performance in large-scale settings. Throughout, the focus is on rigorous theoretical evaluation combined with empirical demonstrations of performance and concordance with the theory.

One topic we consider is modeling the joint distribution of multivariate categorical data, often summarized in a contingency table. Contingency table analysis routinely relies on log-linear models, with latent structure analysis providing a common alternative. Latent structure models lead to a reduced rank tensor factorization of the probability mass function for multivariate categorical data, while log-linear models achieve dimensionality reduction through sparsity. Little is known about the relationship between these notions of dimensionality reduction in the two paradigms. In Chapter 2, we derive several results relating the support of a log-linear model to nonnegative ranks of the associated probability tensor. Motivated by these findings, we propose a new collapsed Tucker class of tensor decompositions, which bridge existing PARAFAC and Tucker decompositions, providing a more flexible framework for parsimoniously characterizing multivariate categorical data. Taking a Bayesian approach to inference, we illustrate empirical advantages of the new decompositions.

Latent class models for the joint distribution of multivariate categorical, such as the PARAFAC decomposition, data play an important role in the analysis of population structure. In this context, the number of latent classes is interpreted as the number of genetically distinct subpopulations of an organism, an important factor in the analysis of evolutionary processes and conservation status. Existing methods focus on point estimates of the number of subpopulations, and lack robust uncertainty quantification. Moreover, whether the number of latent classes in these models is even an identified parameter is an open question. In Chapter 3, we show that when the model is properly specified, the correct number of subpopulations can be recovered almost surely. We then propose an alternative method for estimating the number of latent subpopulations that provides good quantification of uncertainty, and provide a simple procedure for verifying that the proposed method is consistent for the number of subpopulations. The performance of the model in estimating the number of subpopulations and other common population structure inference problems is assessed in simulations and a real data application.

In contingency table analysis, sparse data is frequently encountered for even modest numbers of variables, resulting in non-existence of maximum likelihood estimates. A common solution is to obtain regularized estimates of the parameters of a log-linear model. Bayesian methods provide a coherent approach to regularization, but are often computationally intensive. Conjugate priors ease computational demands, but the conjugate Diaconis--Ylvisaker priors for the parameters of log-linear models do not give rise to closed form credible regions, complicating posterior inference. In Chapter 4 we derive the optimal Gaussian approximation to the posterior for log-linear models with Diaconis--Ylvisaker priors, and provide convergence rate and finite-sample bounds for the Kullback-Leibler divergence between the exact posterior and the optimal Gaussian approximation. We demonstrate empirically in simulations and a real data application that the approximation is highly accurate, even in relatively small samples. The proposed approximation provides a computationally scalable and principled approach to regularized estimation and approximate Bayesian inference for log-linear models.

Another challenging and somewhat non-standard joint modeling problem is inference on tail dependence in stochastic processes. In applications where extreme dependence is of interest, data are almost always time-indexed. Existing methods for inference and modeling in this setting often cluster extreme events or choose window sizes with the goal of preserving temporal information. In Chapter 5, we propose an alternative paradigm for inference on tail dependence in stochastic processes with arbitrary temporal dependence structure in the extremes, based on the idea that the information on strength of tail dependence and the temporal structure in this dependence are both encoded in waiting times between exceedances of high thresholds. We construct a class of time-indexed stochastic processes with tail dependence obtained by endowing the support points in de Haan's spectral representation of max-stable processes with velocities and lifetimes. We extend Smith's model to these max-stable velocity processes and obtain the distribution of waiting times between extreme events at multiple locations. Motivated by this result, a new definition of tail dependence is proposed that is a function of the distribution of waiting times between threshold exceedances, and an inferential framework is constructed for estimating the strength of extremal dependence and quantifying uncertainty in this paradigm. The method is applied to climatological, financial, and electrophysiology data.

The remainder of this thesis focuses on posterior computation by Markov chain Monte Carlo. The Markov Chain Monte Carlo method is the dominant paradigm for posterior computation in Bayesian analysis. It has long been common to control computation time by making approximations to the Markov transition kernel. Comparatively little attention has been paid to convergence and estimation error in these approximating Markov Chains. In Chapter 6, we propose a framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified loss function and computational budget. The results require only ergodicity of the exact kernel and control of the kernel approximation accuracy. The theoretical framework is applied to approximations based on random subsets of data, low-rank approximations of Gaussian processes, and a novel approximating Markov chain for discrete mixture models.

Data augmentation Gibbs samplers are arguably the most popular class of algorithm for approximately sampling from the posterior distribution for the parameters of generalized linear models. The truncated Normal and Polya-Gamma data augmentation samplers are standard examples for probit and logit links, respectively. Motivated by an important problem in quantitative advertising, in Chapter 7 we consider the application of these algorithms to modeling rare events. We show that when the sample size is large but the observed number of successes is small, these data augmentation samplers mix very slowly, with a spectral gap that converges to zero at a rate at least proportional to the reciprocal of the square root of the sample size up to a log factor. In simulation studies, moderate sample sizes result in high autocorrelations and small effective sample sizes. Similar empirical results are observed for related data augmentation samplers for multinomial logit and probit models. When applied to a real quantitative advertising dataset, the data augmentation samplers mix very poorly. Conversely, Hamiltonian Monte Carlo and a type of independence chain Metropolis algorithm show good mixing on the same dataset.

How institutions shaped the last major evolutionary transition to large-scale human societies.

What drove the transition from small-scale human societies centred on kinship and personal exchange, to large-scale societies comprising cooperation and division of labour among untold numbers of unrelated individuals? We propose that the unique human capacity to negotiate institutional rules that coordinate social actions was a key driver of this transition. By creating institutions, humans have been able to move from the default ‘Hobbesian’ rules of the ‘game of life’, determined by physical/environmental constraints, into self-created rules of social organization where cooperation can be individually advantageous even in large groups of unrelated individuals. Examples include rules of food sharing in hunter–gatherers, rules for the usage of irrigation systems in agriculturalists, property rights and systems for sharing reputation between mediaeval traders. Successful institutions create rules of interaction that are self-enforcing, providing direct benefits both to individuals that follow them, and to individuals that sanction rule breakers. Forming institutions requires shared intentionality, language and other cognitive abilities largely absent in other primates. We explain how cooperative breeding likely selected for these abilities early in the Homo lineage. This allowed anatomically modern humans to create institutions that transformed the self-reliance of our primate ancestors into the division of labour of large-scale human social organization.

The Evolution of Massive Young Stellar Objects in the Large Magellanic Cloud

This thesis presents an analysis of the largest catalog to date of infrared spectra of massive young stellar objects in the Large Magellanic Cloud. Evidenced by their very different spectral features, the luminous objects span a range of evolutionary states from those most embedded in their natal molecular material to those that have dissipated and ionized their surroundings to form compact HII regions and photodissociation regions. We quantify the contributions of the various spectral features using the statistical method of principal component analysis. Using this analysis, we classify the YSO spectra into several distinct groups based upon their dominant spectral features: silicate absorption (S Group), silicate absorption and fine-structure line emission (SE), polycyclic aromatic hydrocarbon (PAH) emission (P Group), PAH and fine-structure line emission (PE), and only fine-structure line emission (E). Based upon the relative numbers of sources in each category, we are able to estimate the amount of time massive YSOs spend in each evolutionary stage. We find that approximately 50% of the sources have ionic fine-structure lines, indicating that a compact HII region forms about half-way through the YSO lifetime probed in our study. Of the 277 YSOs we collected spectra for, 41 have ice absorption features, indicating they are surrounded by cold ice-bearing dust particles. We have decomposed the shape of the ice features to probe the composition and thermal history of the ice. We find that most the CO2 ice is embedded a polar ice matrix that has been thermally processed by the embedded YSO. The amount of thermal processing may be correlated with the luminosity of the YSO. Using the Australia Telescope Compact Array, we imaged the dense gas around a subsample of our sources in the HII complexes N44, N105, N113, and N159 using HCO+ and HCN as dense gas tracers. We find that the molecular material in star forming environments is highly clumpy, with clumps that range from subparsec to ~2 parsecs in size and with masses between 10^2 to 10^4 solar masses. We find that there are varying levels of star formation in the clumps, with the lower-mass clumps tending to be without massive YSOs. These YSO-less clumps could either represent an earlier stage of clump to the more massive YSO-bearing ones or clumps that will never form a massive star. Clumps with massive YSOs at their centers have masses larger than those with massive YSOs at their edges, and we suggest that the difference is evolutionary: edge YSO clumps are more advanced than those with YSOs at their centers. Clumps with YSOs at their edges may have had a significant fraction of their mass disrupted or destroyed by the forming massive star. We find that the strength of the silicate absorption seen in YSO IR spectra feature is well-correlated with the on-source HCO+ and HCN flux densities, such that the strength of the feature is indicative of the embeddedness of the YSO. We estimate that ~40% of the entire spectral sample has strong silicate absorption features, implying that the YSOs are embedded in circumstellar material for about 40% of the time probed in our study.

DOC X Just Diary w/o Numbers

On the 15th of April, 1897, a 19 year-old European resident of Baghdad, named Alexander Richard Svoboda, set out on a long journey to Europe by caravan, boat and train. From a large and influential family of merchants, artists, and explorers settled in Ottoman Iraq since the end of the 18th century, Alexander traveled in the company of his parents and a departing British diplomat accompanied by his retinue. They followed a circuitous route through the Middle East to Cairo and thence to Europe on a three and a half month journey which Alexander described day-by-day in a journal written in the Iraqi Arabic of his time.