Scaffold-based molecular design with a graph generative model

Searching for new molecules in areas like drug discovery often starts from the core structures of known molecules. Such a method has called for a strategy of designing derivative compounds retaining a particular scaffold as a substructure. On this account, our present work proposes a graph generative model that targets its use in scaffold-based molecular design. Our model accepts a molecular scaffold as input and extends it by sequentially adding atoms and bonds. The generated molecules are then guaranteed to contain the scaffold with certainty, and their properties can be controlled by conditioning the generation process on desired properties. The learned rule of extending molecules can well generalize to arbitrary kinds of scaffolds, including those unseen during learning. In the conditional generation of molecules, our model can simultaneously control multiple chemical properties despite the search space constrained by fixing the substructure. As a demonstration, we applied our model to designing inhibitors of the epidermal growth factor receptor and show that our model can employ a simple semi-supervised extension to broaden its applicability to situations where only a small amount of data is available.

We propose a scaffold-based graph generative model for designing novel drug candidates that include the desired scaffold as a substructure.     

Cadmium(II) and mercury(II) complexes of an NO2S2-donor macrocycle and its ditopic xylyl-bridged analogue

The NO2S2-donor macrocycle (L1) was synthesised from the ring closure reaction between Boc-N-protected 2,2'-iminobis(ethanethiol) (3) and 2,2'-(ethylenedioxy)bis(benzyl chloride) (4) followed by deprotection of the Boc-group. alpha,alpha'-Dibromo-p-xylene was employed as a dialkylating agent to bridge two L1 to yield the corresponding N-linked product (L2). The X-ray structure of L2 (as its HBr salt) is described. A range of Cd(II) and Hg(II) complexes of L1 (6-9) and L2 (10-12) were prepared and characterised. Reaction of HgX2 (X = Br or I) with L1 afforded [Hg(L1)Br]2[Hg2Br6].2CH2Cl2 6 and [Hg(L1)I(2)] 7, respectively. For 6, the Hg(II) ion in the complex cation has a distorted tetrahedral coordination environment composed of S2N donor atoms from L1 and a bromo ligand. In 7 the coordination geometry is highly distorted tetrahedral, with the macrocycle coordinating in an exodentate manner via one S and one N atom. The remaining two coordination sites are occupied by iodide ions. [Hg(L1)(ClO4)]ClO4 8 was isolated from the reaction of Hg(ClO4)2 and L1. The X-ray structure reveals that all macrocyclic ring donors bind to the central mercury ion in this case, with the latter exhibiting a highly distorted octahedral coordination geometry. The O2S2-donors from the macrocyclic ring define the equatorial plane while the axial positions are occupied by the ring nitrogen as well as by an oxygen from a monodentate perchlorato ion. Reaction of Cd(NO3)(2).4H2O with L1 afforded [Cd(L1)(NO3)2](.)0.5CH2Cl2 9 in which L1 acts as a tridentate ligand, binding exo-fashion via its S2N donors. The remaining coordination positions are filled by two bidentate nitrate ions such that, overall, the cadmium is seven-coordinate. Reactions of HgX2(X = Br or I) with L2 yielded the isostructural 2 : 1 (metal : ligand) complexes, [Hg2(L2)Br4] 10 and [Hg2(L2)I(4)] 11. Each mercury ion has a distorted tetrahedral environment made up of S and N donors from an exodentate L2 and two coordinated halides. Contrasting with this, the reaction of L2 with Cd(NO3)(2).4H2O yielded a 1-D coordination network, {[Cd2(L2)(NO3)4].2CH2Cl2}n 12 in which each ring of L2 is exo-coordinated via two S atoms and one N atom to a cadmium ion which is also bound to one monodentate and one bidentate nitrate anion. The latter also has one of its oxygen atom attached to a neighboring cadmium via a nitroso (mu2-O) bridge such that the overall coordination geometry about each cadmium is seven-coordinate. The [Cd(L2)0.5(NO3)2] units are linked by an inversion to yield the polymeric arrangement.     

Fourier transform infrared spectroscopy study of the effect of pH on anion and cation adsorption onto poly(acrylo‐amidino diethylenediamine)

The role of hydrogen bonding in the chemistry of transition‐metal complexes remains a topic of intense scientific and technological interest. Poly(acrylo‐amidino diethylenediamine) was synthesized to study the effects of hydrogen bonding on complexes at different pHs. The polymer was synthesized through the coupling of diethylene triamine with polyacrylonitrile fiber in the presence of AlCl₃ · 6H₂O addition. The adsorption capacity of this polymer was 11.4 mequiv/g. The ions used for the adsorption test were CrO, PO, Cu²⁺, Ni²⁺, Fe²⁺, and Ag⁺. All experiments were confirmed with Fourier transform infrared. In the study of anion adsorption, at low pHs, only ionic bonds existed, whereas at high pHs, no bonds existed. However, in the middle pH region, both ionic bonds and hydrogen bonds formed between poly(acrylo‐amidino diethylenediamine) and the chromate ion or phosphate ion. When poly(acrylo‐amidino diethylenediamine) and metal ions (Cu²⁺, Ni²⁺, Fe²⁺, and Ag⁺) formed complexes, a hydrogen‐bonding effect was not observed with Fourier transform infrared. The quantity of metal ions adsorbed onto poly(acrylo‐amidino diethylenediamine) followed the order Ag⁺ > Cu²⁺ > Fe²⁺ > Ni²⁺. © 2004 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 42: 2010–2018, 2004     

New Acetylcholinesterase‐Inhibitory Pregnane Glycosides of Cynanchum atratum Roots

Three new pregnane glycosides, cynatroside A ( 1 ), cynatroside B ( 2 ), and cynatroside C ( 3 ), isolated from the roots of Cynanchum atratum (Asclepiadaceae), were characterized as 7β‐{[O‐α‐L ‐cymaropyranosyl‐(1→4)‐O‐β‐D ‐digitoxopyranosyl‐(1→4)‐β‐D ‐oleandropyranosyl]oxy}‐3,4,4a,4b,5,6,7,8,10,10a‐decahydro‐6α‐hydroxy‐4b‐ methyl‐2‐(2‐methyl‐3‐furyl)phenanthren‐1(2H)‐one ( 1 ), 7β‐{[O‐β‐D ‐cymaropyranosyl‐(1→4)‐O‐α‐L ‐diginopyranosyl‐(1→4)‐β‐D ‐cymaropyranosyl]oxy}‐3,4,4a,4b,5,6,7,8,10,10a‐decahydro‐2,6α‐dihydroxy‐4b‐methyl‐2‐(2‐methyl‐3‐furyl)phenanthren‐1(2H)‐one ( 2 ), and 7β‐{[O‐α‐L ‐cymaropyranosyl‐(1→4)‐O‐β‐D ‐digitoxopyranosyl‐(1→4)‐β‐L ‐cymaropyranosyl]oxy}‐3,4,4a,4b,5,6,7,8,10,10a‐decahydro‐2,6α‐dihydroxy‐4b‐methyl‐2‐(2‐methyl‐3‐furyl)phenanthren‐1(2H)‐one ( 3 ), respectively. In addition, ten known constituents were identified, i.e., cynascyroside D ( 4 ), glaucoside C ( 5 ), glaucoside D ( 6 ), atratoside A ( 7 ), 2,4‐dihydroxyacetophenone ( 8 ), 4‐hydroxyacetophenone ( 9 ), syringic acid ( 10 ), azelaic acid ( 11 ), suberic acid ( 12 ), and succinic acid ( 13 ). Among these compounds, 1 – 4 significantly inhibit acetylcholinesterase activity.     

Synthesis and characteristics of novel chelate fiber containing amine and amidine groups

A polyacrylonitrile (PAN) fiber was adopted for the backbone of a chelate polymer and poly(acrylo‐amidino ethylene amine) (PAEA) was prepared through a one‐step reaction between the PAN fiber and ethylenediamine (EDA). The maximum removal capacity and degree of substitution were 7.8 meq per gram of dried PAEA and 98%, respectively. The PAEA was tested as an adsorbent in single and two‐component metal aqueous solutions under changing pH. The Cu²⁺ ion accomplished maximum adsorption amount at pH 3 and the order of maximum adsorbed amounts on PAEA is Cu²⁺ > Ag⁺ > Zn²⁺ > Ni²⁺ > Pb²⁺ in molar basis. FT‐IR spectroscopy was employed to characterize the chemical bonding in metal aqueous solutions and surface morphology was examined using atomic force microscopy (AFM) and scanning electron microscopy (SEM). Copyright © 2004 John Wiley & Sons, Ltd.     

Mechanism and nanosize products of the sol-gel reaction using diphenylsilanediol and 3-methacryloxypropyltrimethoxysilane as precursors

We use a first-principles calculation and small-angle neutron scattering (SANS) to investigate the mechanism and the nanosize products of the sol-gel reaction with diphenylsilanediol (DPD) and 3-methacryloxypropyltrimethoxysilane (MEMO) precursors in synthesizing a hybrid waveguide material. It is predicted that switching between a DPD hydroxyl and a MEMO methoxy with a reaction rate of 6.8 x 10(-6) s(-1) at 300 K is the fastest process for the first reaction step, thus generating diphenylmethoxysilanol (DPM) and 3-methacryloxypropyldimethoxysilanol (MEDO) as products. However, we determine that this reaction pathway could be modified by the presence of the H2O released from a catalyst such as Ba(OH)2.H2O. Next, switching between the DPM hydroxyl and the MEDO methoxy is followed to generate diphenyldimethoxysilane (DPDM) and 3-methacryloxypropylmethoxysilanediol (MEMDO). However, condensation between a MEMDO hydroxyl and a DPDM methoxy is found to be most favorable for the third reaction step, which generates the DPDM-MEMDO dimer and CH3OH molecule as products. In a similar fashion, a DPDM methoxy of the DPDM-MEMDO dimer can condense with a MEMDO hydroxyl of the second DPDM-MEMDO dimer to increase the chain, but its reaction rate of 2.8 x 10(-11) s(-1) is predicted to be about 5 times smaller than that between a DPDM methoxy and a MEMDO hydroxyl. This implies that the reaction rate for the larger nanostructures becomes smaller. Additionally, our SANS measurements determine that the final products from our sol-gel reaction are on the nanometer scale, at sizes from 1.76 to 2.36 nm.     

Preparation of nanoporous MgO-coated TiO2 nanoparticles and their application to the electrode of dye-sensitized solar cells

Sol-gel-derived Mg(OH)(2) gel was coated onto TiO(2) nanoparticles, and the subsequent thermal topotactic decomposition of the gel formed a highly nanoporous MgO crystalline coating. The specific surface area of the electrode that was prepared from the core-shell-structured TiO(2) nanoparticles significantly increased compared with that of the uncoated TiO(2) electrode. The increase in the specific surface area of the MgO-coated TiO(2) electrode was attributed to the highly nanoporous MgO coating layer that resulted from the topotactic reaction. Dye adsorption behavior and solar cell performance were significantly enhanced by employing the MgO-coated TiO(2) electrode. Optimized coating of a MgO layer on TiO(2) nanoparticles enhanced the energy conversion efficiency as much as 45% compared to that of the uncoated TiO(2) electrode. This indicates that controlling the extrinsic parameters such as the specific surface area is very important to improve the energy conversion efficiency of TiO(2)-based solar cells.     

Dissociation of sulfur dioxide by ultraviolet multiphoton absorption between 224 and 232 nm

Multiphoton excitation and dissociation of SO(2) have been investigated in the wavelength range from 224 to 232 nm. Strong evidence is found for two-photon excitation to the H Rydberg state, followed by dissociation to SO + O and ionization of the SO product by absorption of a third photon. The two-photon excitation is resonantly enhanced via the C (1)B(2) intermediate state, and the two-photon yield spectrum thus bears a strong resemblance to the spectrum of this intermediate. Imaging of the O((3)P(2)), S((1)D(2)), and SO products suggests that, following dissociation of SO(2) from the H state, SO is produced in the A and B electronic states. S((1)D(2)) is produced both from two-photon dissociation of SO(2) to give S((1)D(2)) + O(2) and by single-photon dissociation of SO(+). In the former process, the O(2) is likely formed in all of its lowest three electronic states.     

A polymer-based microfluidic device for immunosensing biochips

This paper describes the design, fabrication, and test of a PDMS/PMMA-laminated microfluidic device for an immunosensing biochip. A poly(dimethyl siloxane)(PDMS) top substrate molded by polymer casting and a poly(methyl methacrylate)(PMMA) bottom substrate fabricated by hot embossing are bonded with pressure and hermetically sealed. Two inlet ports and an air vent are opened through the PDMS top substrate, while gold electrodes for electrochemical biosensing are patterned onto the PMMA bottom substrate. The analyte sample is loaded from the sample inlet port to the detection chamber by capillary force, without any external intervening forces. For this and to control the time duration of sample fluid in each compartment of the device, including the inlet port, diffusion barrier, reaction chamber, flow-delay neck, and detection chamber, the fluid conduit has been designed with various geometries of channel width, depth, and shape. Especially, the fluid path has been designed so that the sample flow naturally stops after filling the detection chamber to allow sufficient time for biochemical reaction and subsequent washing steps. As model immunosensing tests for the microfluidic device, functionalizations of ferritin and biotin to the sensing surfaces on gold electrodes and their biospecific interactions with antiferritin antiserum and streptavidin have been investigated. An electrochemical detection method for immunosensing by biocatalyzed precipitation has been developed and applied for signal registration. With the biochip, the whole immunosensing processes could be completed within 30 min.     

Human neural stem cell growth and differentiation in a gradient-generating microfluidic device

This paper describes a gradient-generating microfluidic platform for optimizing proliferation and differentiation of neural stem cells (NSCs) in culture. Microfluidic technology has great potential to improve stem cell (SC) cultures, whose promise in cell-based therapies is limited by the inability to precisely control their behavior in culture. Compared to traditional culture tools, microfluidic platforms should provide much greater control over cell microenvironment and rapid optimization of media composition using relatively small numbers of cells. Our platform exposes cells to a concentration gradient of growth factors under continuous flow, thus minimizing autocrine and paracrine signaling. Human NSCs (hNSCs) from the developing cerebral cortex were cultured for more than 1 week in the microfluidic device while constantly exposed to a continuous gradient of a growth factor (GF) mixture containing epidermal growth factor (EGF), fibroblast growth factor 2 (FGF2) and platelet-derived growth factor (PDGF). Proliferation and differentiation of NSCs into astrocytes were monitored by time-lapse microscopy and immunocytochemistry. The NSCs remained healthy throughout the entire culture period, and importantly, proliferated and differentiated in a graded and proportional fashion that varied directly with GF concentration. These concentration-dependent cellular responses were quantitatively similar to those measured in control chambers built into the device and in parallel cultures using traditional 6-well plates. This gradient-generating microfluidic platform should be useful for a wide range of basic and applied studies on cultured cells, including SCs.