Informática Aplicada a la Investigación Rotating Header Image

Scientific Software

Scientific Software

R eta RStudio

Informazio Orokorra

R 3.3.3 is a freely available language and environment for statistical computing and graphics which provides a wide variety of statistical and graphical techniques: linear and nonlinear modelling, statistical tests, time series analysis, classification, clustering, etc. Please consult the R project homepage for further information.

RStudio and RCommander are a graphical front ends for R.

Instalatutako paketeak

abind, ape, biomformat, cummeRbund, DCGL, DESeq2, DEXSeq, e1071, edgeR, FactoMineR, GEOquery, lavaan, metagenomeSeq, mnormt, optparse, psych, randomForest, Rcmdr, RColorBrewer, ReactomePA, RUVSeq, vegan, WGCNA, xlsx..

Besterenbat behar izanez gero, ezka iezaguzu.

Nola erabili

R exekutatzeko kola sistemaren scriptetan erabili:

/software/bin/R  CMD  BATCH R-input-file.R

eta RStudio erabiltzeko X2Go bitartez egin behar da lehioak ireki ahal izateko, Katramilara edo Txinpartara konektatuz eta exekutatuz:


eta RCommander erabiltzeko X2Go bitartez egin behar da lehioak ireki ahal izateko, Katramilara edo Txinpartara konektatuz eta exekutatu R. Gero R-n kargatu:


Informazio Gehiago

R web orrialdea.

rstudio web orrialdeal.


General information

IDBA-UD 1.1.1 is a iterative De Bruijn Graph De Novo Assembler for Short Reads Sequencing data with Highly Uneven Sequencing Depth. It is an extension of IDBA algorithm. IDBA-UD also iterates from small k to a large k. In each iteration, short and low-depth contigs are removed iteratively with cutoff threshold from low to high to reduce the errors in low-depth and high-depth regions. Paired-end reads are aligned to contigs and assembled locally to generate some missing k-mers in low-depth regions. With these technologies, IDBA-UD can iterate k value of de Bruijn graph to a very large value with less gaps and less branches to form long contigs in both low-depth and high-depth regions.

How to use

To send jobs to the queue you can use the command


which after a few questions configures the job.


IDBA-UD has a good performance and scaling up to 8 cores. Above we did not measure a improvement. In the benchmark the --mimk 40 --step 20 options has been used. When we have decreased the step the the scalling is worse. This trend can be also seen in the second table.

1 core as base 2 cores as base
Cores Time (s) Speed up Performance (%) Speed up Performance (%)
1 480 1 100
2 296 1.6 81 1.0 100
4 188 2.6 64 1.6 79
8 84 5.7 71 3.5 88
12 92 5.2 43 3.2 54

The second benchark has been done with a bigger file with 10 million bases and the  --mink 20 --step 10 --min_support 2 options. We observe a regular behaviour than in the previous benchmark and how the panellization is good up to 4 cores.

Cores Time (s) Speed up Performance
1 13050 1 100
2 6675 2.0 98
4 3849 3.4 85
8 3113 4.2 52
16 2337 5.6 35
20 2409 5.4 27

More information

IDBA-UD web page.


General information

SPAdes 3.6.0 – St. Petersburg genome assembler – is intended for both standard isolates and single-cell MDA bacteria assemblies. It works with Illumina or IonTorrent reads and is capable of providing hybrid assemblies using PacBio, Oxford Nanopore and Sanger reads. You can also provide additional contigs that will be used as long reads. Supports paired-end reads, mate-pairs and unpaired reads. SPAdes can take as input several paired-end and mate-pair libraries simultaneously. Note, that SPAdes was initially designed for small genomes. It was tested on single-cell and standard bacterial and fungal data sets.

How to use

To send jobs to the queue you can use the


command that asks few questions to configure the job.


We have not measure any performance improvement or time reduction when using several cores in a standard calculation like: -pe1-1 file1 -pe1-2 file2 -o outdir

We recommend to use 1 core, unless you know that you can use better performance with several cores.

More information

Web page of SPAdes.


General information

MetAMOS represents a focused effort to create automated, reproducible, traceable assembly & analysis infused with current best practices and state-of-the-art methods. MetAMOS for input can start with next-generation sequencing reads or assemblies, and as output, produces: assembly reports, genomic scaffolds, open-reading frames, variant motifs, taxonomic or functional annotations, Krona charts and HTML report. 1.5rc3 version.

How to use

To send a job to the queue system there is the


command where you answer a few questions to set up the job. Take into account that MetAMOS use a lot of RAM memory, about 1 GB per million reads.

More information

MetAMOS web page.


General information

QIIME (Quantitative Insights Into Microbial Ecology) is an open-source bioinformatics pipeline for performing microbiome analysis from raw DNA sequencing data. QIIME is designed to take users from raw sequencing data generated on the Illumina or other platforms through publication quality graphics and statistics. This includes demultiplexing and quality filtering, OTU picking, taxonomic assignment, and phylogenetic reconstruction, and diversity analyses and visualizations. QIIME has been applied to studies based on billions of sequences from tens of thousands of samples

 How to use

To send QIIME jobs run the command


and answer the questions.


QIIME can use the USEARCH pakage.

More information

QIIME home page.





General Information

Mathematics program that includes symbolic and numeric computing, visualization and programing language. The 10.0 version is installed in Guinness, the 6.0 in the Itanium nodes. Katramila and the newest nodes has the 11.2 version.

Mathematica incluye paralelismo.

Ho to use Mathematica

In interactive mode

The graphical interface can be executed with


and the Mathematica console with


Nota: If you have problems to visualize the fonts maybe you must install them in your local computer.

In the queue system

In the queue scripts use the line

/software/bin/math < input > output

whre input is the file with the Mathematica commands and output is the file where the output will be saved.

More information

Mathematica home page.

Online documentation.


General information

USEARCH is a unique sequence analysis tool that offers search and clustering algorithms that are often orders of magnitude faster than BLAST. We have the free 32 bits version that can not be distributed to third parties and has a 4 GB of RAM limitation.

How to use

To use USEARCH execute


for example

/software/bin/usearch -cluster_otus data.fa -otus otus.fa -uparseout out.up -relabel OTU_ -sizein -sizeout

USEARCH is only available in the xeon20 type nodes.


USEARCH can be use under QIIME.

More information

USEARCH home page.


SAMtools, BCFtools and HTSlib 1.2

General Information

Samtools is a suite of programs for interacting with high-throughput sequencing data. It consists of three separate repositories:

Reading/writing/editing/indexing/viewing SAM/BAM/CRAM format
Reading/writing BCF2/VCF/gVCF files and calling/filtering/summarising SNP and short indel sequence variants
A C library for reading/writing high-throughput sequencing data
Samtools and BCFtools both use HTSlib internally, but these source packages contain their own copies of htslib so they can be built independently.


How to use It

They are installed in /software/samtools-1.2//software/bcftools-1.2/ and  /software/htslib-1.2.1 respectibely.

Something like this should be added in the PBS script.

export PATH=/software/samtools-1.2/bin:/software/bcftools-1.2/bin:$PATH

export LD_LIBRARY_PATH=/software/htslib-1.2.1/lib:$LD_LIBRARY_PATH


More Information


GNU Scientific Libraries for randon number generators, linear algebra, FFT, special functions, etc

To know how to use the


More information at GSL homepage.


General information

2.1.1 release. Trinity, represents a novel method for the efficient and robust de novo reconstruction of transcriptomes from RNA-seq data. Trinity combines three independent software modules: Inchworm, Chrysalis, and Butterfly, applied sequentially to process large volumes of RNA-seq reads. Trinity partitions the sequence data into many individual de Bruijn graphs, each representing the transcriptional complexity at at a given gene or locus, and then processes each graph independently to extract full-length splicing isoforms and to tease apart transcripts derived from paralogous genes. Briefly, the process works like so:

  • Inchworm assembles the RNA-seq data into the unique sequences of transcripts, often generating full-length transcripts for a dominant isoform, but then reports just the unique portions of alternatively spliced transcripts.
  • Chrysalis clusters the Inchworm contigs into clusters and constructs complete de Bruijn graphs for each cluster. Each cluster represents the full transcriptonal complexity for a given gene (or sets of genes that share sequences in common). Chrysalis then partitions the full read set among these disjoint graphs.
  • Butterfly then processes the individual graphs in parallel, tracing the paths that reads and pairs of reads take within the graph, ultimately reporting full-length transcripts for alternatively spliced isoforms, and teasing apart transcripts that corresponds to paralogous genes.

How to use

You can use the


command to submit jobs to the queue system. After answering few questions a script will be created and submitted to the queue system. For advanced users it can be used to generate a sample script.


Trinity can be run in parallel but it is not very efficient above 4 cores with low performance, as can be seen in the the table. Trinity consumes high amounts of RAM.

Performance of Trinity
Cores  1 4 8 12
Time 5189 2116 1754 1852
Speddup 1 2.45 2.96 2.80
Efficiency (%)  100 61 37 23


More information

Trinity web page.


General Information

ABINIT is a package whose main program allows one to find the total energy, charge density and electronic structure of systems made of electrons and nuclei (molecules and periodic solids) within Density Functional Theory (DFT), using pseudopotentials and a planewave or wavelet basis. ABINIT also includes options to optimize the geometry according to the DFT forces and stresses, or to perform molecular dynamics simulations using these forces, or to generate dynamical matrices, Born effective charges, and dielectric tensors, based on Density-Functional Perturbation Theory, and many more properties. Excited states can be computed within the Many-Body Perturbation Theory (the GW approximation and the Bethe-Salpeter equation), and Time-Dependent Density Functional Theory (for molecules). In addition to the main ABINIT code, different utility programs are provided.  netcdf, bigdft, wannier90, etsf_io, libxc, etc. plugins are included. 7.0.4 version is installed.

How to send Abinit

The programs are in /software/bin/abinit. For instance, to execute abinit in the queue system scripts use

/software/bin/abinit/abinit < input.files > out.log

Depending on the number of cores asked to the queue system it will execute in parallel. The jobs can be also submitted with the interactive qsub command.

There are other utilities like aim, anaddb, band2eps, conducti, cut3d, lwf, macroave, mrgddb, mrggkk, newsp, optic, etc.


We have made a small benchmark with the 6.12.3 version.

System 1 cores 4 cores 8 cores
Xeon12 7218 2129 1045
Xeon8 7275 1918 1165
Itanium 8018 2082 1491
Opteron 17162 4877 2802

We observe that Abinit works fine in itanium and xeon nodes while in opteron nodes is slower.

More information

Abinit web page.

Online manuals.


General Information

1.3.2 version of ABySS (Assembly By Short Sequences). ABySS is a de novo, parallel, paired-end sequence assembler that is designed for short reads. ABySS can be executed in parallel.

See also the installed velvet and comparing both we have published article.

How to use

The executables can be found in /software/abyss/bin. To run abyss in a script type in it:

/software/abyss/bin/abyss-pe [abyss-pe options]


See also the installed velvet and comparing both we have published article.


Some benchmarks has been performed with ABySS. They have been performed using file from an Illumina HiSeq2000 NGS with 100 bp per sequence. In the table 1 we can see an example about how ABySS scales as a function of the number of cores. As we can see ABySS scales very up to 8 cores. The results is valid unless for more than 10e6 sequences.

Table 1. Execution time of abyss-pe in seconds as a function the number of cores
cores 2 4 8 12 24
Time (s) 47798 27852 16874 14591 18633
Aceleration 1 1.7 2.8 3.3 2.6
Performance(%) 100 86 71 55 21

Execution time

We have analized as well the execution time as a function of the size of the data. In the table 2 we  observe how from 1 million to 10 millions of sequences the execution time increases by 10 as well. From 10 to 100 millions of sequences the time increases a little more, between 10 t0 20. Therefore, the behavior is more or less lineal.

Table 2. Execution time in seconds of abyss-pe executed in 2, 4 y 8 cores as a function of the number of processed sequences.
sequences 10e6 10e7 10e8
Time in 2 cores (s) 247 2620 47798
Time in 4 cores (s) 134 1437 27852
Time in 8 cores (s) 103 923 16874

RAM memory

In these kind of programs more important than the execution time, which is reasonable, is the RAM memory usage, which can limit the calculation type. In the table 3 we observe how the RAM increases as a function of the number of sequences. We also show the logarithms of the measured values which has been used for a lineal regression. The jobs has been performed in 12 cores.

Tabl3 3. RAM memory used by abyss-pe as a function of the number of processed sequences. The logarithms of the measured values are also shown.
sequences 10e6 5*10e6 10e7 5*10e7 10e8
RAM (GB) 4.0 7.6 11 29 44
log(sequences) 6 6.7 7 7.7 8
log(RAM) 0.60 0.88 1.03 1.46 1.65

From the values of the table we obtain a fitting of the RAM in GB as a function of the number of sequences (s) to the equation


o equivalently



The memory usage is smaller than in other assemblers like Velvet, see as well the report Velvet performance in the machines of the Computing Service of the UPV/EHU and comparing both we have published article. In addition, the parallelization with MPI of ABySS allows to aggregate the RAM memory of several nodes to perform larger calculations.

More information

ABySS web page.
Velvet assembler.
Velvet performance in the machines of the Computing Service of the UPV/EHU report.

Velvet and ABySS performance in the machines of the Computing Service of the UPV/EHU, post in the hpc blog.


Clean reads

General information

0.2.2 Version. clean_reads cleans NGS (Sanger, 454, Illumina and solid) reads. It can trim:

  • bad quality regions
  • adaptors
  • vectors
  • regular expresssions

It also filters out the reads that do not meet a minimum quality criteria based on the sequence length and the mean quality. It can run in parallel.

Ho to use

To submit clean_reads jobs to the queue system execute the command


It will ask few questions to build the script and submit it to the queue.


clean_reads can be executed in parallel and scales well up to 8 cores. For 12 cores the performance is very poor. In the table 1 we show the results of the benchmark. They have been executed in a 12 cores node with E5645 Xeon processors.

Execution time in seconds as a function of the number of cores
cores 1 4 8 12
Time (s) 1600 422 246 238
Speedup 1 3.8 6.5 6.7
Performance (%) 100 95 81 56

The used command has been

clean_reads -i in.fastq -o out.fastq -p illumina -f fastq -g fastq -a a.fna -d UniVec -n 20 --qual_threshold=20 --only_3_end False -m 60 -t 12

More information

clean_reads web page.


General information

1.2.03 version. Velvet is a set of algorithms manipulating de Bruijn graphs for genomic and de novo transcriptomic Sequence assembly. It was designed for short read sequencing technologies, such as Solexa or 454 Sequencing and was developed by Daniel Zerbino and Ewan Birney at the European Bioinformatics Institute. The tool takes in short read sequences, removes errors then produces high quality unique contigs. It then uses paired-end read and long read information, when available, to retrieve the repeated areas between contigs.

See also the installed ABySS and comparing both we have published article.

How to use

To run velveth or velvetg add in your scripts for the Torque queue system the corresponding command:

/software/bin/velvet/velveth [velvet options]
/software/bin/velvet/velvetg [velvet option]


Velvet has been compiled with parallel support througth OpenMP. We have measured the perfomance and the results are available in the report about the Velvet performance in the machines of the Computing Service of the UPV/EHU. Velvet uses huge amount of RAM for large calculations and we have measured it. In the report some simple formulas are obtained to predict the use of RAM for their input files, so the researches can know the needed RAM before start the calculations and in this way can plan their research.

See also the installed ABySS and comparing both we have published article.

More information

Velvet web page.
Velvet performance in the machines of the Computing Service of the UPV/EHU report.
Velvet and ABySS performance in the machines of the Computing Service of the UPV/EHU, post in the hpc blog.


General Information

2.2.24 version of BLAST de NCBI. Due to performance reasons it has not been installed in Itanium nodes.


Data bases

The Service has installed several data bases, contact the technicians to use them or install new ones.

How to use

To submit jobs to the queue system we strongly recomend to use the command


it will make some questions to prepare the job

Performance and gpuBLAST

We have compared BLAST with mpiBLAST and gpuBLAST, the result of the bechmarks are in the blog of Service. mpiBLAST is installed in the Service.


More information

BLAST web page.

Blast2GO y mpiBLAST is also installed.


General information

4.1 version.

Genepop is a population genetics software package, which has options for the following analysis: Hardy Weinberg equilibrium, Linkage Disequilibrium, Population Differentiation, Effective number of migrants, Fst or other correlations.

How to use

To execute Genepop in the queue system you must include in the script of the queue system:

/software/bin/Genepop < input_file

where input_file has the options for Genepop, i.e., the answer to Genepop when it runs in interactive mode. We recommend to use qsub in interactive mode to submit the jobs


More information

Genepop web page.


General information

1.1.3 version. CLUMPP is a program that deals with label switching and multimodality problems in population-genetic cluster analyses. CLUMPP permutes the clusters output by independent runs of clustering programs such as structure, so that they match up as closely as possible. The user has the option of choosing one of three algorithms for aligning replicates, with a tradeoff of speed and similarity to the optimal alignment.

How to use

To execute CLUMPP in the queue system you must include in the script of the queue system:


with the corresponding options of structure. We recommend to use qsub in interactive mode to submit the jobs


More information

CLUMPP web page.


General information

2.33 version.

The program structure is a free software package for using multi-locus genotype data to investigate population structure. Its uses include inferring the presence of distinct populations, assigning individuals to populations, studying hybrid zones, identifying migrants and admixed individuals, and estimating population allele frequencies in situations where many individuals are migrants or admixed. It can be applied to most of the commonly-used genetic markers, including SNPS, microsatellites, RFLPs and AFLPs.

How to use

To execute the graphical user interface execute in Péndulo, Maiz or Guinness


To execute graphical applications read how to connect to Arina.

To execute structure in the queue system you must include in the script of the queue system:


with the corresponding options of structure. We recommend to use qsub in interactive mode to submit the jobs


More information

Structure web page.

MCCCS Towhee 7.0.2

Towhee is a Monte Carlo molecular simulation code originally designed for the prediction of fluid phase equilibria using atom-based force fields and the Gibbs ensemble with particular attention paid to algorithms addressing molecule conformation sampling. The code has subsequently been extended to several ensembles, many different force fields, and solid (or at least porous) phases.

General Information

Towhee serves as a useful tool for the molecular simulation community and allows science to move forward more quickly by eliminating the need for individual research groups to rewrite routines that already exist and instead allows them to focus on algorithm advancement, force field development, and application to interesting systems.

Towhee may use different type of ensembles and Monte Carlo moves implemented into Towhee and can alos used different  force fields included with the distribution. (See here for more information )

How to Use


  • To send Towhee to the queue system  use the send_gulp utility. When executed,
    shows the command syntax, which is summarized below:
  • send_towhee JOBNAME NODES PROCS_PER_NODE TIME [ MEM ] [``Other queue options'' ]
JOBNAME: Is the  name of the Output.
NODES: Number of nodes.
PROCS: Number of  processors.
TIME: Time requested to the queue system, format hh:mm:ss.
MEM: Optional. Memory in Gb ( It will used 1GB/core if not set).
[``Other Torque Options'' ] Optional. There is the possibility to pass more variables to the queuing system.
See examples below.   More information about this options


We send a Towhee job1 to 1 node, 4 processors on that node, with a requested time of 4 hours . The results will be in the OUT file.

send_towhee OUT 1 4 04:00:00

We send job2  to 2 compuation nodes, 8 processors on each node, with a requested time of 192 hours, 8 GB of RAM and to start running after work 1234.arinab is finished:

send_towhee OUT 2 8 192:00:00 8 ``-W depend=afterany:1234'

We  send the input job3 to 4 nodes and 4 processors on each node, with arequested time of 200:00:00 hours, 2 GB of RAM and we request to be send an email at the beginning and end of the calculation to the direction specified.

send_towhee OUT 4 4 200:00:00 2 ``-m be -M''

send_towhee command copies the contents of the directory from which the job is sent to /scratch or / gscratch, if we use 2 or more nodes. And there is where the calculation is done.

Jobs Monitoring

To facilitate monitoring and/or control of the Towhee calculations, you can use remote_vi

remote_vi JOBID

It show us the *.out file  (only if it was sent using send_towhee).

More information


General information

XCrySDen is a crystalline and molecular structure visualisation program, which aims at display of isosurfaces and contours, which can be superimposed on crystalline structures and interactively rotated and manipulated.

How to use

To use XCrySDen execute:


More information

XCrySDen web orrialdea.


General information

Gretl (Gnu Regression, Econometrics and Time-series Library) is a software for econometric analysis. 1.9.6 version is installed.


  • Incluye una gran variedad de estimadores: mínimos cuadrados, máxima verosimilitud, GMM; de una sola ecuación y de sistemas de ecuaciones
  • Métodos de series temporales: ARMA, GARCH, VARs y VECMs, contrastes de raíces unitarias y de cointegración, etc.
  • Variables dependientes limitadas: logit, probit, tobit, regresión por intervalos, modelos para datos de conteo y de duración, etc.
  • Los resultados de los modelos se pueden guardar como ficheros LaTeX, en formato tabular y/o de ecuación.
  • Incluye un lenguaje de programación vía ‘scripts’ (guiones de instrucciones): las órdenes se pueden introducir por medio de los menús o por medio de guiones.
  • Estructura de bucles de instrucciones para simulaciones de Monte Carlo y procedimientos de estimación iterativos.
  • Controlador gráfico mediante menús, para el ajuste fino de los gráficos Gnuplot.
  • Enlace a GNU R, GNU Octave y Ox para análisis más sofisticados de los datos.

How to use

In order to use Gretl execute


More information

Gretl web page.


General information

IBM ILOG CPLEX Optimizer’s mathematical optimization technology enables smarter decision-making for efficient resource utilization. CPLEX provides robust algorithms for demanding problems: IBM ILOG CPLEX Optimizer has solved problems with millions of constraints and variables. 12.6.3 version.


  • Automatic and dynamic algorithm parameter control
    IBM ILOG CPLEX Optimizer automatically determines “smart” settings for a wide range of algorithm parameters, usually resulting in optimal linear programming solution performance. However, for a more hands-on approach, dozens of parameters may be manually adjusted, including algorithmic strategy controls, output information controls, optimization duration limits, and numerical tolerances.
  • Fast, automatic restarts from an advanced basis
    Large problems can be modified, and then solved again in a fraction of the original solution time.
  • A variety of problem modification options, such as:
    – The ability to add and delete variables
    – The ability to add and delete constraints
    – The ability to modify objective, right-hand side, bound and matrix coefficients
    – The ability to change constraint types
  • A wide variety of input/output options, such as:
    – Problem files: read/write MPS files, IBM ILOG CPLEX Optimizer LP files, MPS basis and revise files, binary problem/basis files
    – Log files: session information and various solution reports
    – Solution files: ASCII and binary solution files
    – IBM ILOG CPLEX Optimizer messages: Each message type (such as RESULTS, WARNINGS or ERRORS) can be directed to specified files, or completely suppressed.
  • Post solution information and analysis, including:
    – Objective function value
    – Solution variable and slack values
    – Constraint dual values (shadow prices)
    – Variable reduced costs
    – Right-hand side, objective function, and bound sensitivity ranges
    – Basic variables and constraints
    – Solution infeasibilities (if any exist)
    – Iteration/node count, solution time, process data
    – Infeasibility (IIS) finder for diagnosing problem infeasibilities
    – Feasibility optimizer for automatic correction of infeasible models

How to use

To use CPLEX execute:



There is a small benchmark usingCOIN-OR, see COIN-OR web page.

More information

CPLEX web page.


General information

1.7.5 version. The Computational Infrastructure for Operations Research (COIN-OR**, or simply COIN) project is an initiative to spur the development of open-source software for the operations research community. It has been konpiled using CPLEX.

How to use

The blis, cbc, clp, OSSolverService and symphony executables are installed in /software/bin/CoinAll. To use, for example, clp execute in the torque scripts:


More information

Coin-or home page.


General information

4.02 version of the MD program for macromolecules, polymers, ionic systems, solutions and other molecular systems. Developed at the Daresbury Laboratory. In Pendulo the 2.2 version remains. There is already the DL_POLY_CLASSIC version which currently is not been developed.

How to submit to the queue

The program is installed in all the architectures, Arina and Pendulo (DL_POLY 2.2). To execute it include in the scripts:


The program will exekute in GPGPUs if it starts in these kind of nodes. Besides, they can be selected by using the gpu label within the queue system.

The GUI is also installed. To execute it use:


Some utilities has been installed in the /software/bin/DL_POLY/ directory.


We show a small benchmarks performed with dl_ploly_4.02. We stady the parallelization as well as the performance of the GPGPUs.

System 1 cores 4 cores 8 cores 16 cores 32 cores 64 cores
Itanium 1.6 GHz 1500 419 248 149 92 61
Opteron 1230 503 264 166 74
Xeon 2.27 GHz 807 227 126 67 37 25

We show in the firs benchamrk that DL_POLY scales very well and that the xeon nodes are the fastest ones, so we recomend them for large jobs.

System 1 cores 2 cores 4 cores 8 cores 16 cores 32 cores
Itanium 1.6 GHz 2137 303 165 93 47
Opteron 1592 482 177 134 55
Xeon 2.27 GHz 848 180 92 48 28
1 GPGPU 125 114 104 102
2 GPGPU 77 72 69
4 GPGPU 53 50
8 GPGPU 37
System 1 cores 2 cores 4 cores 8 cores 16 cores 32 cores 64 cores
Xeon 2.27 GHz 2918 774 411 223 122 71
1 GPGPU 362 333 338 337
2 GPGPU 240 222 220
4 GPGPU 145 142
8 GPGPU 97

We show that the GPGPUs speedup the calculation but each time we double the number of GPGPUs the speed up is multiplied but only 1.5. Because of this for large number of GPGPUs or cores is better to use the paralelization over cores. For example, one node has 8 cores and 2 GPGPUS. The 2 GPGPUs need 220 s while 8 cores need 411 s. Still 4 GPGPUs are faster than 16 cores but 32 cores with 71 s are faster than 8 GPGPUs that need 97 s. Therefore, the GPGPUS can speedup jobs in PCs or single nodes, but for jobs that require higher parallelization the cores parallelization is more effective.

DL_POLY is designed for big systems and the use up to thousand of cores. According to the documentation:

The DL_POLY_4 parallel performance and efficiency are considered very-good-to-excellent as long as (i) all CPU cores are loaded
with no less than 500 particles each and (ii) the major linked cells algorithm has no dimension less than 4.

More information

DL_POLY web page.

DL_POLY user guide (pdf).

DL_POLY GUI user guide (pdf).


General information

Version 7.9 (2009b) in the opteron nodes and 7.12 (2010a) in the Xeon nodes. Mathematical program that includes  high-level technical computing language and interactive environment for algorithm development, data visualization, data analysis, and numeric computation.

How to use

Intereactively (with graphical interface):

Matlab is only available for Opteron and Xeon so have to be used interactively from Guinness or Maiz. The graphical interface can be executed with:


In the queue system

Add the following line in the Torque scripts

/software/bin/matlab < input_file > output_file

where input_file is the file with the matlab commands and output_file is the standar output file name.

With old versions we have had problems with .m files with more than 169 lines. If this happens call your .m program from another small .m program.

To submit the calculation to the nodes add the xeon or the opteron label to submit the calculation to these kink of nodes. In addition, we only have licences to use 2 nodes, so we have the “matlab” special queue to manage them. You must use it. Therefore, your script must contain

#PBS -q matlab
#PBS -l nodes=1:ppn=1:xeon

or for the opteron nodes:

#PBS -q matlab
#PBS -l nodes=1:ppn=1:opteron


Some benchmarks have been realized comparing Octave, Scilab y Matlab.


Ezyfit tool box is installed.

Optimization tool box is installed.

Statistics tool box is installed.

More information

Matlab home page.

Online documentation.


General information

Amsterdam Density Functional (2016.01)

The ADF package is software for first-principles electronic structure calculations. ADF is used by academic and industrial researchers in such diverse fields as pharmacochemistry and materials science. It is particularly popular in the research areas of homogeneous and heterogeneous catalysis, inorganic chemistry, heavy element chemistry, various types of spectroscopy, and biochemistry. Key Benefits of ADF

  • Spectroscopic properties and environments for any type of molecule
  • Excels in transition and heavy metal compounds
  • Accurate, robust, and fast
  • Expert staff and active community
  • Uses Slater functions, beats Gaussians!

ADF can be used for finite systems (molecules) and periodic systems by using ADFBAND. ADF includes graphical interfaces and analysis tools available from Katramila server.

The Service has license for 4 cores.

How to use

To execute the graphical interfaces for input and result visualization use in Katramila:


To submit an ADF input name created with adfinput program we recommend to use the send_adf


<code>send_adf JOBNAME NODES PROCS[property] TIME MEM [``Otherqueue options'' ]
  • JOBNAME: Input name without the extension
  • NODES: Number of nodes
  • PROCS: Number of processors.
  • TIME: Requested time in hh:mm:ss format.
  • MEM: Memory in GB.
  • [“Otras opciones de Torque”] Other queue oprions.


To submit the input file to 4 processors, with a 20 hours walltime and 4 GB of RAM use

send_adf 1 4 20:00:00 4

More information

ADF web page.

ADF documentation.



General information

2018 version. GROMACS is a versatile package to perform molecular dynamics, i.e. simulate the Newtonian equations of motion for systems with hundreds to millions of particles.

It is primarily designed for biochemical molecules like proteins, lipids and nucleic acids that have a lot of complicated bonded interactions, but since GROMACS is extremely fast at calculating the nonbonded interactions (that usually dominate simulations) many groups are also using it for research on non-biological systems, e.g. polymers.

How to use


To send gromacs to the queue system  use the send_gmx utility. When executed, shows the command syntax, which is summarized below:

send_gmx ``JOB and Options'' NODES PROCS_PER_NODE TIME MEM [``Other queue options'']
``JOB and Options'': options for the calculation and input of GROMACS name extension. It is very important to keep the quotes.
NODES: Number of nodes.
PROCS: Number of  processors.
TIME: Time requested to the queue system, format hh:mm:ss.
MEM: Memory in Gb.
[``Otras opciones de Torque'' ] There is the possibility to pass more variables to the queuing system.
See examples below.   More information about this options


We send the gromacs input job1 to 1 node, 4 processors on that node, with a requested time of 4 hours and 1 GB of RAM:

send_gmx ``-s job1.tpr'' 1 4 04:00:00 1

We send job2  to 2 compuation nodes, 8 processors on each node, with a requested time of 192 hours, 8 GB of RAM and to start running after work 1234.arinab is finished:

send_gmx ``-s job2.tpr'' 2 8 192:00:00 8 ``-W depend=afterany:1234'

We  send the input job3 to 4 nodes and 4 processors on each node, with a requested time of 200:00:00 hours, 2 GB of RAM and we request to be send an email at the  beginning and end of the calculation to the direction specified.

send_gmx ``-s  job.tpr'' 4 4 200:00:00 2 ``-m be -M''

send_gmx command copies the contents of the directory from which the job is sent to /scratch or /gscratch, if we use 2 or more nodes. And there is where the calculation is done.

Jobs Monitoring

To facilitate monitoring and/or control of the gromacs calculations, you can use remote_vi which shows the md.log file (only if it was sent using send_gmx).

More information

Scientific software

Quantum Mechanics

Abinit ok ok
ADF 2017.110 2017.110 2017.110
BigDFT ok ok
Casino 2.4 ok
Dirac 08 ok
Espresso ok ok
Gamess Jan 2009 ok
Gaussian 03 & 09 ok ok ok
Jaguar ok ok ok
MIKA .81 ok
NBO 5 ok
NWChem 6.3 ok ok
Orca 3.03  ok ok
PSI4 ok ok
Qbox ok ok
Qsite  ok ok ok
Siesta 2.0.1 ok
Siesta 3.0 ok
TB-LMTO 4.6 ok
Terachem ok
Turbomole 7.0  ok ok
VASP ok ok ok
Wien2K ok ok
Yambo 3.2.2 ok ok

BioChemistry / Molecular Mechanics

Amber ok ok
DL_POLY ok ok
Gromacs ok ok
Gulp 4.0 ok
Macromodel ok ok ok
NAMD 2.6 ok
Towhee 7.0.2   ok


Coin-or ok ok
CPLEX ok ok
Grace 5.1.19 ok ok
Mathematica  8.0 ok ok
Matlab ok ok
Matplotlib ok
Octave 3.2.3 ok
Octave 3.2.4 ok
R, RCommander and RStudio 3.3.3 3.3.2 2.7.0
Scilab 5.1.1 ok
Scilab 5.2.2 ok


ABySS ok
BEAST ok ok
Blast2Go ok
Clean_reads ok
Cufflinks ok
Genepop ok
IDBA-UD ok ok
MetAMOS ok
mpiBLAST ok
SPAdes ok ok
Structure ok
TopHat ok
Trinity ok
Velvet ok


Gaussview ok
Maestro ok ok
Molden ok ok ok
NX client ok
x2Go ok ok
XCrysDen ok
Xmakemol  ok ok
VMD ok ok

Other Software

GAP 4.4 ok
Gretl ok
HTK 3.4.1
OOMMF  ok ok
STAR-CCM+ ok ok ok
WRF 3.1 ok



General information

Maestro is the unified interface for all Schrödinger software. Impressive rendering capabilities, a powerful selection of analysis tools, and an easy-to-use design combine to make Maestro a versatile modeling environment for all researchers.

In Ginness and Péndulo servers 9.0 version is installed, Maiz has the 8.5 version. We recommend to use Guinness to run Maestro, is the newest server and which support lastest versions.

How to use

Maestro is going to use graphical applications so the researcher must ensure that can open this windows in his computer. How to do this is explained in the access guide of the service.

To execute Maestro run in the terminal


More information

Maestro web page.

Schrödinger web page.


General information

TeraChem 1.45 is general purpose quantum chemistry software designed to run on NVIDIA GPU architectures under a 64-bit Linux operating system. Some of TeraChem features include:

  • Full support for both NVIDIA Tesla and Fermi GPUs
  • Restricted Hartree-Fock and Kohn-Sham single point energy and gradient calculations
  • Various DFT functionals (BLYP, B3LYP, PBE, etc) and DFT grids (800-80,000 grid points per atom)
  • Empirical dispersion corrections (DFT-D)
  • Geometry optimization and transition state search (including constraints)
  • Ab initio molecular dynamics (NVE and NVT ensembles)
  • Support of multiple-GPU systems
  • Up to 1000 times faster than conventional CPU-based codes
  • Designed for large molecules – reads/writes PDB files directly
  • Optimization including geometric constraints
  • Improved mixed-precision for increased accuracy

How to send Terachem

send_terachem command exist to facilitate the job submision to the cluster.

More information

Terachem web page.

User guide.