EDISON-WMW: Exact Dynamic Programing Solution of the Wilcoxon–Mann–Whitney Test

In many research disciplines, hypothesis tests are applied to evaluate whether findings are statistically significant or could be explained by chance. The Wilcoxon–Mann–Whitney(WMW) test is among the most popular hypothesis tests in medicine and life science to analyze if two groups of samples are equally distributed. This nonparametric statistical homogeneity test is commonly applied in molecular diagnosis. Generally, the solution of the WMW test takes a high combinatorial effort for large sample cohorts containing a significant number of ties. Hence, P value is frequently approximated by a normal distribution. We developed EDISON-WMW, a new approach to calculate the exact permutation of the two-tailed unpaired WMW test without any corrections required and allowing for ties. The method relies on dynamic programing to solve the combinatorial problem of the WMW test efficiently. Beyond a straightforward implementation of the algorithm, we presented different optimization strategies and developed a parallel solution. Using our program,the exact P value for large cohorts containing more than 1000 samples with ties can be calculated within minutes. We demonstrate the performance of this novel approach on randomly-generated data, benchmark it against 13 other commonly-applied approaches and moreover evaluate molecular biomarkers for lung carcinoma and chronic obstructive pulmonary disease(COPD). We foundthat approximated P values were generally higher than the exact solution provided by EDISONWMW. Importantly, the algorithm can also be applied to high-throughput omics datasets, where hundreds or thousands of features are included. To provide easy access to the multi-threaded version of EDISON-WMW, a web-based solution of our algorithm is freely available at http://www.ccb.uni-saarland.de/software/wtest/.     

Comparison of batch and continuous multi‐column protein A capture processes by optimal design

Multi‐column capture processes show several advantages compared to batch capture. It is however not evident how many columns one should use exactly. To investigate this issue, twin‐column CaptureSMB, 3‐ and 4‐column periodic counter‐current chromatography (PCC) and single column batch capture are numerically optimized and compared in terms of process performance for capturing a monoclonal antibody using protein A chromatography. Optimization is carried out with respect to productivity and capacity utilization (amount of product loaded per cycle compared to the maximum amount possible), while keeping yield and purity constant. For a wide range of process parameters, all three multi‐column processes show similar maximum capacity utilization and performed significantly better than batch. When maximizing productivity, the CaptureSMB process shows optimal performance, except at high feed titers, where batch chromatography can reach higher productivity values than the multi‐column processes due to the complete decoupling of the loading and elution steps, albeit at a large cost in terms of capacity utilization. In terms of trade‐off, i.e. how much the capacity utilization decreases with increasing productivity, CaptureSMB is optimal for low and high feed titers, whereas the 3‐column process is optimal in an intermediate region. Using these findings, the most suitable process can be chosen for different production scenarios.     

A novel algorithm with differential evolution and coral reef optimization for extreme learning machine training

Extreme learning machine (ELM) is a novel and fast learning method to train single layer feed-forward networks. However due to the demand for larger number of hidden neurons, the prediction speed of ELM is not fast enough. An evolutionary based ELM with differential evolution (DE) has been proposed to reduce the prediction time of original ELM. But it may still get stuck at local optima. In this paper, a novel algorithm hybridizing DE and metaheuristic coral reef optimization (CRO), which is called differential evolution coral reef optimization (DECRO), is proposed to balance the explorative power and exploitive power to reach better performance. The thought and the implement of DECRO algorithm are discussed in this article with detail. DE, CRO and DECRO are applied to ELM training respectively. Experimental results show that DECRO-ELM can reduce the prediction time of original ELM, and obtain better performance for training ELM than both DE and CRO.     

An improved neutral landscape model for recreating real landscapes and generating landscape series for spatial ecological simulations

Many studies have assessed the effect of landscape patterns on spatial ecological processes by simulating these processes in computer‐generated landscapes with varying composition and configuration. To generate such landscapes, various neutral landscape models have been developed. However, the limited set of landscape‐level pattern variables included in these models is often inadequate to generate landscapes that reflect real landscapes. In order to achieve more flexibility and variability in the generated landscapes patterns, a more complete set of class‐ and patch‐level pattern variables should be implemented in these models. These enhancements have been implemented in Landscape Generator (LG), which is a software that uses optimization algorithms to generate landscapes that match user‐defined target values. Developed for participatory spatial planning at small scale, we enhanced the usability of LG and demonstrated how it can be used for larger scale ecological studies. First, we used LG to recreate landscape patterns from a real landscape (i.e., a mountainous region in Switzerland). Second, we generated landscape series with incrementally changing pattern variables, which could be used in ecological simulation studies. We found that LG was able to recreate landscape patterns that approximate those of real landscapes. Furthermore, we successfully generated landscape series that would not have been possible with traditional neutral landscape models. LG is a promising novel approach for generating neutral landscapes and enables testing of new hypotheses regarding the influence of landscape patterns on ecological processes. LG is freely available online.     

GAP promoter‐based fed‐batch production of highly bioactive core streptavidin by Pichia pastoris

Streptavidin is a homotetrameric protein binding the vitamin biotin and peptide analogues with an extremely high affinity, which leads to a large variety of applications. The biotin‐auxotrophic yeast Pichia pastoris has recently been identified as a suitable host for the expression of the streptavidin gene, allowing both high product concentrations and productivities. However, so far only methanol‐based expression systems have been applied, bringing about increased oxygen demand, strong heat evolution and high requirements for process safety, causing increased cost. Moreover, common methanol‐based processes lead to large proportions of biotin‐blocked binding sites of streptavidin due to biotin‐supplemented media. Targeting these problems, this paper provides strategies for the methanol‐free production of highly bioactive core streptavidin by P. pastoris under control of the constitutive GAP promoter. Complex were superior to synthetic production media regarding the proportion of biotin‐blocked streptavidin. The optimized, easily scalable fed‐batch process led to a tetrameric product concentration of up to 4.16 ± 0.11 µM of biotin‐free streptavidin and a productivity of 57.8 nM h^?1 based on constant glucose feeding and a successive shift of temperature and pH throughout the cultivation, surpassing the concentration in un‐optimized conditions by a factor of 3.4. Parameter estimation indicates that the optimized conditions caused a strongly increased accumulation of product at diminishing specific growth rates (μ ≈ D < 0.01 h^?1), supporting the strategy of feeding. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:855–864, 2016     

Toward intensifying design of experiments in upstream bioprocess development: An industrial Escherichia coli feasibility study

In this study, step variations in temperature, pH, and carbon substrate feeding rate were performed within five high cell density Escherichia coli fermentations to assess whether intraexperiment step changes, can principally be used to exploit the process operation space in a design of experiment manner. A dynamic process modeling approach was adopted to determine parameter interactions. A bioreactor model was integrated with an artificial neural network that describes biomass and product formation rates as function of varied fed‐batch fermentation conditions for heterologous protein production. A model reliability measure was introduced to assess in which process region the model can be expected to predict process states accurately. It was found that the model could accurately predict process states of multiple fermentations performed at fixed conditions within the determined validity domain. The results suggest that intraexperimental variations of process conditions could be used to reduce the number of experiments by a factor, which in limit would be equivalent to the number of intraexperimental variations per experiment. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:1343–1352, 2016     

A comparison of four different conformations adopted by human telomeric G‐quadruplex using computer simulations

The telomeric G‐quadruplexes for their unique structural features are considered as potential anticancer drug targets. These, however, exhibit structural polymorphism as different topology types for the intra‐molecular G‐quadruplexes from human telomeric G‐rich sequences have been reported based on NMR spectroscopy and X‐ray crystallography. These techniques provide detailed atomic‐level information about the molecule but relative conformational stability of the different topologies remains unsolved. Therefore, to understand the conformational preference, we have carried out quantum chemical calculations on G‐quartets; used all‐atom molecular dynamics (MD) simulations and steered molecular dynamics (SMD) simulations to characterize the four human telomeric G‐quadruplex topologies based on its G‐tetrad core‐types, viz., parallel, anti‐parallel, mixed‐(3 + 1)‐form1 and mixed‐(3 + 1)‐form2. We have also studied a non‐telomeric sequence along with these telomeric forms giving a comparison between the two G‐rich forms. The structural properties such as base pairing, stacking geometry and backbone conformations have been analyzed. The quantum calculations indicate that presence of a sodium ion inside the G‐tetrad plane or two potassium ions on both sides of the plane give it an overall planarity which is much needed for good stacking to form a helix. MD simulations indicate that capping of the G‐tetrad core by the TTA loops keep the terminal guanine bases away from water. The SMD simulations along with equilibrium MD studies indicate that the parallel and non‐telomeric forms are comparatively less stable. We could come to the conclusion that the anti‐parallel form and also the mixed‐(3 + 1)‐form1 topology are most likely to represent the major conformation., 2016. © 2015 Wiley Periodicals, Inc. Biopolymers 105: 83–99, 2016     

Overexpression of Candida rugosa lipase Lip1 via combined strategies in Pichia pastoris

In this study, combined strategies were employed to heterologously overexpress Candida rugosa lipase Lip1 (CRL1) in a Pichia pastoris system. The LIP1 gene was systematically codon-optimized and synthesized in vitro. The Lip1 activity of a recombinant strain harboring three copies of the codon-optimized LIP1 gene reached 1200 U/mL in a shake flask culture. Higher lipase activity, 1450 U/mL, was obtained using a five copy number construct. Co-expressing one copy of the ERO1p and BiP chaperones with Lip1p, the CRL1 lipase yield further reached 1758 U/mL, which was significantly higher than that achieved by expressing Lip1p alone or only co-expressing one molecular chaperone. When cultivated in a 3 L fermenter under optimal conditions, the recombinant strain GS115/87-ZA-ERO1p-BiP #7, expressing the molecular chaperones Ero1p and BiP, produced 13,490 U/mL of lipase activity at 130 h, which was greater than the 11,400 U/mL of activity for the recombinant strain GS115/pAO815-α-mCRL1 #87, which did not express a molecular chaperone. This study indicates that a strategy of combining codon optimization with co-expression of molecular chaperones has great potential for the industrial-scale production of pure CRL1.