积分java源码-compchem_start:如何开始计算化学研究。针对我们实验室的新人，一般来说可能有用

# Computational chemistry - how to get started This document has resources for people new to computational quantum chemistry. Mainly directed for people starting at [our lab](https://blogs.helsinki.fi/johansson-group/). This list is currently quite rough looking, I'm constantly making updates :) List of abbreviations: * DFT - Density functional theory * QC - Quantum chemistry * QM - Quantum mechanics * MD - Molecular dynamics # Books and lecture notes ## Quantum chemistry * [Atkins: Molecular Quantum Mechanics](https://www.amazon.com/Molecular-Quantum-Mechanics-Peter-Atkins/dp/0199541426) * [Piela: Ideas of Quantum Chemistry](https://www.amazon.com/Ideas-Quantum-Chemistry-Lucjan-Piela/dp/0444522271) ## Computational chemistry ### General * [Szabo, Ostlund: Modern Quantum Chemistry](https://www.amazon.com/Modern-Quantum-Chemistry-Introduction-Electronic/dp/0486691861) * Quite dense * [Jensen: Introduction to Computational Chemistry](https://www.amazon.com/Introduction-Computational-Chemistry-Frank-Jensen/dp/1118825993/ref=dp_ob_image_bk) * Accessible * [Bachrach: Computational Organic Chemistry](https://www.amazon.com/Computational-Organic-Chemistry-Steven-Bachrach/dp/1118291921/ref=dp_ob_title_bk) * Nice examples, easy to read * [Cramer: Essential of Computational Chemistry](https://www.amazon.com/Essentials-Computational-Chemistry-Theories-Models/dp/0470091827) * Nice explanations, accessible * [Helgaker, Jorgensen, Olsen: Molecular Electronic-Structure Theory](https://www.amazon.com/Molecular-Electronic-Structure-Theory-Trygve-Helgaker/dp/0470017600/ref=dp_ob_title_bk) * Very dense, a reference book ### DFT * [Koch, Holthausen: A Chemist's Guide to Density Functional Theory ](https://www.amazon.com/Chemists-Guide-Density-Functional-Theory/dp/3527303723) ### Other * [Reiher: Relativistic Quantum Chemistry](https://www.amazon.com/Relativistic-Quantum-Chemistry-Fundamental-Molecular/dp/3527334157) * [Berendsen: Simulating the Physical World](https://www.amazon.com/Simulating-Physical-World-Hierarchical-Mechanics/dp/0521835275) * Simulating materials, describes different models needed for different scales very well # Working with remote clusters * All the heavy computational work is done on remote computers * All these computers run on linux * DigitalOcean (provides virtual private servers for websites etc) has good tutorials: [link](https://www.digitalocean.com/community/tutorials/) * Terminal basics * Learn to use shell efficiently * mkdir, mv, cp, ls, pwd, ... * [Basic Linux Navigation and File Management](https://www.digitalocean.com/community/tutorials/basic-linux-navigation-and-file-management) * Connect with ssh if you use linux/mac, with [putty](https://putty.org/) etc on Windows * [Tutorial ](https://www.digitalocean.com/community/tutorials/ssh-essentials-working-with-ssh-servers-clients-and-keys) * Copying files done with SCP * [Tutorial](https://www.digitalocean.com/community/questions/how-to-copy-files-from-one-server-to-another-droplet) ## CSC * The calculations require a lot of resources * [CSC](https://research.csc.fi/guides) provides computational resources for us * Taito supercluster * Sisu supercomputer * A lot of chemistry software preinstalled * [Listed here](https://research.csc.fi/software?p_p_id=122&p_r_p_564233524_categoryId=53920) * Access to [several databases](https://research.csc.fi/chemistry/databases) * [Cambridge Structural Database System](https://research.csc.fi/-/csd) * X-ray structures of molecules, useful toolkits for searching structures * Web interface [WebCSD](https://www.ccdc.cam.ac.uk/structures/) * Search with DOI etc * The .cif files can be opened with avogadro, chemcraft, ... ## Resource allocation * Clusters have a lot of users, workload manager is used to allocate resources * Define the resources you need * Amount of cores, memory, time, ... * [SLURM](https://slurm.schedmd.com/) workload manager is used in our environments * Batch job file * Used to define the calculation * Resources and software needed * [CSCs tutorial](https://research.csc.fi/taito-constructing-a-batch-job-file) * Here's another[ tutorial](https://github.com/abalter/slurm-tutorial/wiki/slurm-tutorial) # Used methods ## Quantum chemistry $H\Psi = E\Psi$ * Basis sets * BO approximation * HF * Electron correlation * DFT * Functionals ## Workflow 1) Build the molecule with some visual tool (see below) * [xyz file format](https://en.wikipedia.org/wiki/XYZ_file_format) 2) Create an input file * Each QC program has their own 3) Send the job to queue 4) Check results * Errors are usually due to * faulty input file * not enough resources, such as time, memory * convergence problems * several ways to proceed, see Jensen's book 5) Collect and analyze results * Spreadsheets, Python, ... ## Molecular dynamics * Classical simulations * Potential is defined as a sum of bond, angle, dihedral, van der Waals, electrostatic terms * Parameters from experiments, QC calculations * No quantum mechanical effects * Bond breaking, polarization, ... * Not very useful in our stuff, we mostly look at small molecules # Software ## Quantum chemistry * [ORCA](https://orcaforum.cec.mpg.de/) * [On CSC](https://research.csc.fi/software-details/-/asset_publisher/bwNv9EAV4eYX/content/orca) * Active community, good forum * [ORCA input library](https://sites.google.com/site/orcainputlibrary/) is *amazing*. Easy to get started with this! * Free for scientists! :) * Good for spectroscopy * DLPNO-CCSD(T) method! * [Turbomole](http://www.turbomole.com/) * [On CSC](https://research.csc.fi/-/turbomole?redirect=https%3A%2F%2Fresearch.csc.fi%2Fsoftware%3Fp_p_id%3D101_INSTANCE_wfvLxzjnZlJx%26p_p_lifecycle%3D0%26p_p_state%3Dnormal%26p_p_mode%3Dview%26p_p_col_id%3Dcolumn-2%26p_p_col_pos%3D1%26p_p_col_count%3D4%26p_r_p_564233524_categoryId%3D61753%26p_r_p_564233524_resetCur%3Dtrue) * Fast! * No typical input files * Jobs are set up using the interactive define-script * Not very user friendly, must be scripted away when running large-scale calculations * [Gaussian](http://gaussian.com/) * [On CSC](https://research.csc.fi/software-details/-/asset_publisher/bwNv9EAV4eYX/content/gaussian) * [NWChem](http://www.nwchem-sw.org/index.php/Main_Page) * [On CSC](https://research.csc.fi/software-details/-/asset_publisher/bwNv9EAV4eYX/content/nwchem?redirect=https%3A%2F%2Fresearch.csc.fi%2Fsoftware-details%3Fp_p_id%3D101_INSTANCE_bwNv9EAV4eYX%26p_p_lifecycle%3D0%26p_p_state%3Dnormal%26p_p_mode%3Dview%26p_p_col_id%3Dcolumn-2%26p_p_col_count%3D1) * Open-source! :) * Good for LARGE problems, scales to thousands of processors ## Molecular dynamics * [GROMACS](http://www.gromacs.org/) * The manual is a good intro to MD! ## Visualization and * [Avogadro](https://avogadro.cc/) * Open source, quite powerful * Best tool for building molecules, UFF force field for preoptimization * Conformation search * Visualization of orbitals, vibrations, Bader analysis, ... * [VMD](http://www.ks.uiuc.edu/Research/vmd/) * Excellent and scriptable program, heavily used in MD field * [Chemcraft](https://www.chemcraftprog.com/) * Good for many manipulations, visualization * Not free, 150 days free trial period ## Wavefunction analysis * [MultiWFN](http://sobereva.com/multiwfn/) * Excellent tool for wave function analysis * Pretty much everything is implemented * Takes as an input eg. a molden-file * [NCIPlot](http://www.lct.jussieu.fr/pagesperso/contrera/nciplot.html) * Visual analysis of weak interactions ## Scripting and working with data ### Python * Best choice in my opinion! * I use it for everything: scientific computing, workflow automation, web development, deep learning, home automation... * One of the most popular languages * Excellent re