Until such a time that computers are powerful enough to routinely perform ab initio simulation of large biomolecules, there will remain a demand for less expensive computational tools. Classical force field methods are widely used for the simulation of large molecules. However, their low computational cost comes at the price of introducing approximations to the description of the system, for example atomic point charges and Hooke type potentials. The quantum chemical topological force field, QCTFF, removes the classical approximations and uses a machine learning method, kriging, to build models that map ab initio atomic properties to changes in the internal coordinates of a chemical system. The atomic properties come from quantum chemical topology, QCT, and include atomic multipole moments and also energy terms from the interacting quantum atoms (IQA) energy decomposition scheme. By using atomic multipole moments, the electrostatic interactions between atoms is described in a more rigorous fashion than most classical force fields, and polarisation is captured through the use of kriging models. In this thesis, the QCTFF approach has been applied to a selection of test cases including small molecular dimers and amino acids. Kriging models are built using a “training set” of molecular geometries, and an investigation of different approaches for sampling amino acids is provided. The concept of the “atomic horizon sphere” is discussed, where the effect on the multipole moments of an atom in an increasingly large environment is investigated. This is an important investigation required to guide the development of future QCTFF training sets. Investigations into the effect of deprotonation of basic and acidic amino acids side chains is provided, as well as a study of the short range repulsion between atoms.

The assessment of the lung weight in vivo is possible with the quantitative computer tomography (qCT) analysis. Especially in acute lung injury (ALI) the knowledge of the lung weight can help to identify the etiology of lung-dysfunction. The current definition of ALI is orientating on parameters such as impaired oxygenation or radiological opacifications. With this definition a heterogeneous group is captured. There might be dysfunction of the lung due to atelectasis or due to edema, both leading to impaired oxygenation. For the clinician it is important to differentiate between atelectasis and edema. For example, in patients with edematous lungs the clinician is focusing on prevention of secondary lung injury whereas in atelectasis the clinician is targeting a more aggressive treatment. The method of qCT has the potential to differentiate atelectasis from edema and could thus provide valuable information for managing trauma patients fulfilling commonly used criteria for ALI. So far a reference value for normal lung weights in ventilated patients is not available. In recent studies the lung weights of ALI-Patients where compared to healthy spontaneous breathing patients. Effects of a positive end expiratory pressure ventilation (PEEP) or possible influence of intravenous fluid substitution were ignored. The aim of this work was define a reference value of normal lung-weights in ventilated trauma patients. To reveal possible effects of PEEP or intravenous fluid substitution on the lung weight we provided a comparison group of spontaneous breathing trauma patients. In this prospective observational study CTs of trauma patients with normal lungs who underwent emergency CT were selected and two subgroups formed for spontaneous breathing (n = 31) and mechanically ventilated patients (n = 44). The decision whether a lung was normal was based on independent reviewers of the CT images. The arterial partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2) had to be greater than 400mmHg. Demographic data, ventilation and clinical parameters of each patient where obtained from the patient data management system. In demographic data mechanically ventilated patients did not differ from the spontaneous breathing patients (only significant variation in sex, with a higher male proportion in the ventilated group). Mechanically ventilated patients were ventilated with PEEP of 10 mmHg at the time of the CT acquisition. The PaO2/FiO2 ratio was 550  74 mmHg. Mechanically ventilated patients received significant more intravenous fluid substitution (p = 0.02). The lung weight in mechanically ventilated patients was 873  124 g Standard deviation (SD), in spontaneous breathing patients 866  169 g SD. The validity of our method was reviewed by placing a water filled plastic bottle next to the thorax. The mass was calculated in two ways: by quantitative computed tomography and by the volumetric mass density of water. A deviation of 2% could be shown Our results suggest that lung weights of mechanically ventilated patients with normal lungs do not differ from those of other with normal lungs (Gattinoni 2006, Puybasset 2000). In conclusion, a moderate PEEP neither a moderate intravenous fluid substitution do not affect the lung weight. The lung weights assessed in this work can be used as reference values, especially, for the group of the trauma-associated ALI. With these results it is possible to identify pathological lung weights. Furthermore it gives a tool in identifying the etiology of ALI and therefore it helps the clinician in making the right therapeutic decisions.

Typical, potential-driven force fields have been usefully applied to small molecules for decades. However, complex effects such as polarisation, π systems and hydrogen bonding remain difficult to model while these effects become increasingly relevant. In fact, these complex electronic effects become crucial when considering larger biological molecules in solution. Instead, machine learning can be used to recognise patterns in chemical behaviour and predict them, sacrificing computational efficiency for accuracy and completeness of the force field. The kriging machine learning method is capable of taking the geometric features of a molecule and predicting its electrostatic properties after being trained using ab initio data of the same system. We present significant improvements in functionality, application and understanding of the kriging machine learning as part of an electrostatic force field. These improvements are presented alongside an up-scaling of the problems the force field is applied to. The force field predicts electrostatic energies for all common amino acids with a mean error of 4.2 kJmol-1 (1 kcal mol-1), cholesterol with a mean error of 3.9 kJmol-1 and a 10-alanine helix with a mean error of 6.4 kJmol-1. The kriging machine learning has been shown to work identically with charged systems, π systems and hydrogen bonded systems. This work details how different chemical environments and parameters affect the kriging model quality and assesses optimal methods for computationally-efficient kriging of multipole moments. In addition to this, the kriging models have been used to predict moments for atoms they have had no training data for with little loss in accuracy. Thus, the kriging machine learning has been shown to produce transferable models.

To better define the relationship between vascular calcification and bone mass/structure, we assessed abdominal aortic calcification (AAC), BMD, and bone microstructure in an age-stratified, random sample of 693 Rochester, MN, residents. Participants underwent QCT of the spine and hip and high-resolution pQCT (HRpQCT) of the radius to define volumetric BMD (vBMD) and microstructural parameters. AAC was quantified with the Agatston scoring method. In men, AAC correlated with lower vertebral trabecular and femoral neck vBMD (p < 0.001), but not after age or multivariable (age, body mass index, smoking status) adjustment. Separation into <50 and ≥50 yr showed this pattern only in the older men. BV/TV and Tb.Th inversely correlated with AAC in all men (p < 0.001), and Tb.Th remained significantly correlated after age adjustment (p < 0.05). Tb.N positively correlated with AAC in younger men (p < 0.001) but negatively correlated in older men (p < 0.001). The opposite was true with Tb.Sp (p = 0.01 and p < 0.001, respectively). Lower Tb.N and higher Tb.Sp correlated with AAC in older men even after multivariable adjustment. Among all women and postmenopausal women, AAC correlated with lower vertebral and femoral neck vBMD (p < 0.001) but not after adjustment. Lower BV/TV and Tb.Th correlated with AAC (p = 0.03 and p = 0.04, respectively) in women, but not after adjustment. Our findings support an age-dependent association between AAC and vBMD. We also found that AAC correlates with specific bone microstructural parameters in older men, suggesting a possible common pathogenesis for vascular calcification and deterioration in bone structure. However, sex-specific differences exist.

Ένα από τα βασικά χαρακτηριστικά της ομόζυγης β-θαλασσαιμίας είναι η οστεοπάθεια, η οποία αποτελεί μία πολυπαραγοντική διαταραχή, που δεν έχει πλήρως διευκρινισθεί. Μελετήσαμε τους οσφυϊκούς σπονδύλους 48 ασθενών με τις μεθόδους Dual-Energy X-ray Absorptiometry (DXA) και Quantitative Computed Tomography (QCT), και εστιάσαμε στις δομικές οστικές ιδιότητες, όπως προσδιορίζονται από την υψηλής ευκρίνειας Υπολογιστική τομογραφία (HRCT). Οι τιμές της οστικής πυκνότητας (BMD values) εκφράσθηκαν ως Z-scores και τα αποτελέσματα συσχετίσθηκαν. Εκτιμήθηκε η επίδραση της ηλικίας, του φύλου, του τύπου της θαλασσαιμίας και των ορμονικών παραγόντων στις τιμές ΒΜD. Αξιολογήσαμε, με βάση την HRCT, την ακεραιότητα του φλοιού και τον αριθμό και πάχος των δοκίδων της σπογγώδους ουσίας. Με βάση τον αριθμό των δοκίδων ταξινομήσαμε τους ασθενείς σε κλίμακα τριών βαθμίδων. Τα αποτελέσματά μας έδειξαν ότι ο συνολικός επιπολασμός της οστεοπόρωσης με την μέθοδο DXA ήταν 44 % και με την QCT 6 %. Και οι δύο μέθοδοι έδειξαν μία αρνητική συσχέτιση μεταξύ της ηλικίας και της BMD, ενώ οι ορμονικοί παράγοντες παρουσίασαν συσχετίσεις τόσο με τις μετρήσεις της QCT όσο και με τις αντίστοιχες της DXA. Ο συντελεστής συσχέτισης μεταξύ της BMD της DXA και της σπογγώδους BMD της QCT ήταν 0,545 (p<0,001) ενώ η αντίστοιχη τιμή για τα Ζ-scores ήταν 0,491 (p<0,001). Η ομαδοποίηση των ασθενών σε φυσιολογικούς, οστεοπενικούς και οστεοπορωτικούς, με βάση το Ζ της QCT, ήταν σε καλύτερη συμφωνία με την ταξινόμηση με βάση τον αριθμό των δοκίδων (K=0,209, p=0,053), σε σύγκριση με την ομαδοποίηση σύμφωνα με το Ζ της μεθόδου DXA (K=0,145, p=0,120). Η εκτίμηση του φλοιού με την HRCT έδειξε διακοπές στη συνέχειά του σε 15 ασθενείς. Και οι δύο μέθοδοι δείχνουν μία επιδείνωση της οστεοπόρωσης με την πρόοδο της ηλικίας. Η ανεπάρκεια των ορμονών συσχετίζεται με την θαλασσαιμική οστεοπόρωση, ενώ η οτπική εκτίμηση του φλοιώδους οστού δείχνει ότι οι ενδιάμεσου τύπου θαλασσαιμικοί πάσχουν σε μεγαλύτερο βαθμό από τους ασθενείς με μείζονα μορφή θαλασσαιμίας. Με τον αριθμό των δοκίδων ως δείκτη οστεοπόρωσης, φαίνεται ότι η QCT μπορεί να εκτιμήσει την οστεοπάθεια καλύτερα από την DXA. Δεδομένου ότι η QCT έχει την ικανότητα να μετρήσει την οστική πυκνότητα του σπογγώδους και φλοιώδους οστού, ξεχωριστά, μπορεί να παρέχει πρώιμη ένδειξη του ποιο από τα δύο μεταβάλλεται πιο γρήγορα και σε τι βαθμό. - / Osteopathy, as a major feature of homozygous beta-thalassaemia, is a multifactorial disorder, not fully understood. We studied the lumbar vertebrae of 48 patients using Dual-Energy X-ray Absorptiometry (DXA) and Quantitative Computed Tomography (QCT), and we focused on structural properties, assessed by High Resolution Computed Tomography (HRCT). Bone Mineral Density (BMD) values were expressed as Z scores and the results were correlated. The effect of age, sex, type of thalassaemia and hormonal factors on BMD was assessed. We estimated, with HRCT, the cortex integrity and the number and thickness of trabeculae; the latter were classified to a three-grade scale. Our results showed the overall prevalence of osteoporosis to be 44 % with DXA and 6 % with QCT. Both techniques revealed an inverse correlation between age and BMD, whereas hormonal factors demonstrated associations with QCT and DXA measurements. The correlation coefficient between DXA’s BMD and QCT’s trabecular BMD was 0.545 (p<0.001) whereas the corresponding value for Z scores was r=0.491 (p<0.001). The classification of the patients into normal, osteopenic and osteoporotic categories, using QCT’s Z, was in better agreement with the assignment based on trabecular number (K=0.209, p=0.053) than the classification using DXA’s Z (K=0.145, p=0.120). Cortex evaluation by HRCT showed discontinuity in 15 patients. Both methods indicate a progression of osteoporosis with age. Hormonal deficiency is associated with thalassaemic osteoporosis whereas the visual estimation of cortex indicate that TI could be more affected than TM. Using the trabecular number as an indicator of osteoporosis, it seems that QCT may evaluate osteopathy better than DXA. Since the former has the ability to measure trabecular and cortical BMD separately, it could give early indication of which changes more rapidly and to what degree.

Muscle strength is assumed to be closely related with BMD, the so called determinant of bone strength, however, new methods for bone strength measurement are arising. The purpose of this study was to determine the relationship between bone mineral density (BMD), muscle strength and natural frequency of the radius in the dominant and non-dominant arm in healthy and osteopenic individuals aged between 50-70 years. Sixty sedentary male (thirty healthy and thirty osteopenic) participated this study. Bone mineral density assessment was performed by dual x-ray absorbtiometry (DEXA) and quantitative computed tomography (QCT), whereas muscle strength was measured by an isokinetic dynamometer quantitatively. Natural frequency of the radius was determined by a dual channel frequency analyzer. Differences between BMD, muscle strength and natural frequency in healthy and osteopenic participants according to dominancy were examined by Analysis of Variance (ANOVA). Pearson Product Correlation Coefficient test was conducted to determine the magnitude of the correlation between cortical, trabecular and average BMD, muscle strength and natural frequency. Results demonstrated a statistically significant difference between BMD, natural frequency and muscle strength in the dominant arm of both groups. There was also a significant difference in the non-dominant arm in terms of BMD, natural frequency and muscle strength, except in total work in the non-dominant arms. Moreover, there was a moderate positive correlation between BMD measured by DEXA and natural frequency in the dominant arm (r = ,59 / p &lt / .001) and non-dominant arm (r = 0,64 / p &lt / 0.001), whereas the muscle strength was correlated to BMD with a low positive correlation in terms of peak torque in extension (r = ,36 / p = ,005), peak torque in flexion (r = ,31 / p = ,016), total work in extension (r = ,28 / p = ,030) and total work in flexion (r = ,27 / p = ,041) in the dominant arms. The correlation between muscle strength and BMD was not significant in the non-dominant arm. The highest correlation between natural frequency and bone geometry parameters was observed in cortical thickness (r = ,82 / p = ,02). A statistically significant positive correlation (r = ,81 / p = ,04) was also observed between average BMD measured by QCT and by DEXA. In summary, according to the findings of this study, it can be concluded that vibration analysis is a precise method in predicting bone strength that depends highly on its size, shape and the distribution of its trabecular and cortical components.

The assessment of the lung weight in vivo is possible with the quantitative computer tomography (qCT) analysis. Especially in acute lung injury (ALI) the knowledge of the lung weight can help to identify the etiology of lung-dysfunction. The current definition of ALI is orientating on parameters such as impaired oxygenation or radiological opacifications. With this definition a heterogeneous group is captured. There might be dysfunction of the lung due to atelectasis or due to edema, both leading to impaired oxygenation. For the clinician it is important to differentiate between atelectasis and edema. For example, in patients with edematous lungs the clinician is focusing on prevention of secondary lung injury whereas in atelectasis the clinician is targeting a more aggressive treatment. The method of qCT has the potential to differentiate atelectasis from edema and could thus provide valuable information for managing trauma patients fulfilling commonly used criteria for ALI. So far a reference value for normal lung weights in ventilated patients is not available. In recent studies the lung weights of ALI-Patients where compared to healthy spontaneous breathing patients. Effects of a positive end expiratory pressure ventilation (PEEP) or possible influence of intravenous fluid substitution were ignored. The aim of this work was define a reference value of normal lung-weights in ventilated trauma patients. To reveal possible effects of PEEP or intravenous fluid substitution on the lung weight we provided a comparison group of spontaneous breathing trauma patients. In this prospective observational study CTs of trauma patients with normal lungs who underwent emergency CT were selected and two subgroups formed for spontaneous breathing (n = 31) and mechanically ventilated patients (n = 44). The decision whether a lung was normal was based on independent reviewers of the CT images. The arterial partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2) had to be greater than 400mmHg. Demographic data, ventilation and clinical parameters of each patient where obtained from the patient data management system. In demographic data mechanically ventilated patients did not differ from the spontaneous breathing patients (only significant variation in sex, with a higher male proportion in the ventilated group). Mechanically ventilated patients were ventilated with PEEP of 10 mmHg at the time of the CT acquisition. The PaO2/FiO2 ratio was 550  74 mmHg. Mechanically ventilated patients received significant more intravenous fluid substitution (p = 0.02). The lung weight in mechanically ventilated patients was 873  124 g Standard deviation (SD), in spontaneous breathing patients 866  169 g SD. The validity of our method was reviewed by placing a water filled plastic bottle next to the thorax. The mass was calculated in two ways: by quantitative computed tomography and by the volumetric mass density of water. A deviation of 2% could be shown Our results suggest that lung weights of mechanically ventilated patients with normal lungs do not differ from those of other with normal lungs (Gattinoni 2006, Puybasset 2000). In conclusion, a moderate PEEP neither a moderate intravenous fluid substitution do not affect the lung weight. The lung weights assessed in this work can be used as reference values, especially, for the group of the trauma-associated ALI. With these results it is possible to identify pathological lung weights. Furthermore it gives a tool in identifying the etiology of ALI and therefore it helps the clinician in making the right therapeutic decisions.:Bibliografische Beschreibung 3 Abkürzungsverzeichnis 4 Einleitung 6 Gewichtsbestimmungen in der Medizin 6 Quantitative Computertomographie (qCT): Volumen- und Massebestimmung 7 Klinische Anwendung von Analysen des Lungengewichtes 9 Zielsetzung 11 Patienten und Methoden 12 Maschinell beatmete Patienten (maschinell B.) 12 Spontan atmende Patienten (spontan A.) 14 CT-Untersuchung 14 Quantitative CT-Analyse: Segmentierung der CT-Bilder 15 Validierung der Methodik 18 Auswertung der ROI 18 Statistische Analyse 19 Ergebnisse 21 Demographische Daten 21 Ergebnisse der Lungengewichtsanalyse 25 Ergebnisse der Validierung 28 Diskussion 30 Ausblick 40 Zusammenfassung / Abstract 41 Literaturverzeichnis 46

Yes / The optimisation of a peptide-capped glycine using the novel force field FFLUX is presented. FFLUX is a force field based on the machine-learning method kriging and the topological energy partitioning method called Interacting Quantum Atoms. FFLUX has a completely different architecture to that of traditional force fields, avoiding (harmonic) potentials for bonded, valence and torsion angles. In this study, FFLUX performs an optimisation on a glycine molecule and successfully recovers the target density-functional theory energy with an error of 0.89 ± 0.03 kJ mol−1. It also recovers the structure of the global minimum with a root-mean-squared deviation of 0.05 Å (excluding hydrogen atoms). We also show that the geometry of the intra-molecular hydrogen bond in glycine is recovered accurately. / EPSRC Established Career Fellowship [grant number EP/K005472]

No description available.

Force fields have been an integral part of computational chemistry for decades, providing invaluable insight and facilitating the better understanding of biomolecular system behaviour. Despite the many benefits of a force field, there continue to be deficiencies as a result of the classical architecture they are based upon. Some deficiencies, such as a point charge electrostatic description instead of a multipole moment description, have been addressed over time, permitted by the ever-increasing computational power available. However, whilst incorporating such significant improvements has improved force field accuracy, many still fail to describe several chemical effects including polarisation, non-covalent interactions and secondary/tertiary structural effects. Furthermore, force fields often fail to provide consistency when compared with other force fields. In other words, no force field is reliably performing more accurately than others, when applied to a variety of related problems. The work presented herein develops a next-generation force field entitled FFLUX, which features a novel architecture very different to any other force field. FFLUX is designed to capture the relationship between geometry and energy through a machine learning method known as kriging. Instead of a series of parameterised potentials, FFLUX uses a collection of atomic energy kriging models to make energy predictions. The energies describing atoms within FFLUX are obtained from the Interacting Quantum Atoms (IQA) energy partitioning approach, which in turn derives the energies from the electron density and nuclear charges of topological atoms described by Quantum Chemical Topology (QCT). IQA energies are shown to provide a unique insight into the relationship between geometry and energy, allowing the identification of explicit atoms and energies contributing towards torsional barriers within various systems. The IQA energies can be modelled to within 2.6% accuracy, as shown for a series of small systems including weakly bound complexes. The energies also allow an interpretation of how an atom feels its surrounding environment through intra-atomic, covalent and electrostatic energetic descriptions, which typically are seen to converge within a ~7 - 8 A horizon radius around an atom or small system. These energy convergence results are particularly relevant to tackling the transferability theme within force field development. Where energies are seen to converge, a proximity limit on the geometrical description needed for a transferable energy model is defined. Finally, the FFLUX force field is validated through successfully optimising distorted geometries of a series of small molecules, to near-ab initio accuracy.