Frequently Asked Questions


  1. What are the major differences between the two databases listed here?
  2. How do you define experimentally confirmed genes?
  3. What are the sources of BacMet data?
  4. How to download the 'Experimentally Confirmed' (Bacmet_EXP) and the BacMet 'Predicted' (BacMet_ALL) databases?
  5. Which database and what criteria should I use in my work to search for resistance genes in my NGS dataset?
  6. How can I save my search results from web-interface?
  7. How to search biocide resistance genes in NGS data using BacMet database?
  8. What are the software requirements for BacMet-Scan?
  9. How do I install BacMet-Scan?
  10. Can I use BacMet-Scan on Windows?
  11. Can I use BacMet-Scan on MacOS X?
  12. BacMet-Scan does not work, and gives an error message beginning with "ERROR! Could not locate BLAST binaries!"
  13. I have a really large data set. BacMet-Scan seems to be using BLAST for matching, is any faster tools supported by BacMet-Scan?
  14. Both BLAST and BLAT have multi-threaded implementations. Are those supported by BacMet-Scan?
  15. I have a set of predicted ORFs, is it possible to scan those with BacMet-Scan?
  16. I want to scan multiple datasets with BacMet-Scan and compare the results. Is there some convenient output option for that?
  17. I want to see the actual BLAST (or BLAT/vmatch) report. Is that possible?
  18. There seem to be a lot of different output options; can I get all of them in some convenient way?
  19. Does the database give a PubMed reference for each gene?
  20. How to cite BacMet database?
  21. Who should I contact, if I have any questions about the database?


  1. What are the major differences between the two databases listed here?
  2. In this web-server, we have listed two databases called as 'BacMet Experimenatlly Confirmed database (BacMet_EXP)' and the 'BacMet Predicted database (BacMet_ALL)'. The major differences are:

    'BacMet Experimentally Confirmed database' contains genes that are experimentally verified to provide resistance, whereas 'BacMet Predicted Database' contains the same genes as well as genes with similar sequences, predicted to have a resistance function. The predicted resistance genes have been collected using similarity searches in the NCBI non-redundant database, and are believed to be representative of the diverse range of resistance genes available in bacterial communities. An individual similarity cut-off has been applied to chromosomal genes, while a fixed cut-off has been applied to genes that have been found on plasmids (please see also question 5)

    The genes in the experimentally confirmed database are biased towards model bacteria, since those are highly over-represented in microbiology laboratory experiments (e.g. Escherichia coli, Pseudomonas aeruginosa etc.)

  3. How do you define experimentally confirmed genes?
  4. A gene has been considered experimentally confirmed only if the removal or insertion/overexpression of the gene resulted in a decreased or increased phenotypic resistance, respectively.

  5. What are the sources of BacMet data?
  6. Data on source organism, location (plasmid/chromosomal), gene description etc. were collected from publicly available resources such as NCBI protein database, UniprotKB database, and Transporter Classification Database (TCDB) and then manually scrutinized before included into the experimentally confirmed database. On the other hand, all the data in the BacMet Predicted database were collected from only NCBI protein database.

  7. How to download the 'Experimentally Confirmed' (Bacmet_EXP) and the BacMet 'Predicted' (BacMet_ALL) databases?
  8. All the datasets are available to download from the 'Download' section. We have provided a tutorial on how to download data in the 'Tutorials' section.

  9. Which database and what criteria should I use in my work to search for resistance genes in my NGS dataset?
  10. If the main purpose of a study is to identify potential risk for dissemination of resistance between bacteria, a database of only mobile or acquired (i.e. horizontally transferable) resistance genes is often desired. Resistance genes may differ between species and/or occur in different forms that are not (yet) experimentally investigated. For plasmid-borne resistance genes, this might be less of an issue since mobile resistance genes are often highly conserved in different species. However, if we screen genomes or metagenomes against a database of only experimentally verified sequences using strict criteria, we might miss many resistance genes that are present on chromosomes in different species. On the other hand, if we relax the search criteria, we might detect false-positives that have a conserved section (e.g. a domain) similar to the resistance gene but have a different function than conferring resistance. For each of the chromosomal genes with experimentally confirmed resistance function, we have therefore studied the functional annotation of similar genes. This annotation has been used to set individually adapted similarity cut-offs to reduce the probability of including genes with non-resistance function in the predicted database, at the same time not setting the cut-off too high, which would exclude many resistance genes. Thus, rather than using the experimentally confirmed database for screening of metagenomes and applying a uniform, more relaxed cut-off criteria, we recommend to use predicted database with a criterion of about 85-90% sequence identity with the full length coverage of the short reads (75-300+ bps) against resistance genes. Similarly, the predicted database is useful for investigating the presence of resistance genes in genomes of species that are not closely related to the ones our experimental knowledge about the genes resistance function are derived from. For more details see Pal et al. 2017 (page 24; http://hdl.handle.net/2077/48671)

  11. How can I save my search results from web-interface?
  12. You can 'select' and 'copy' the html table output from the search results and then 'paste' into a new document (e.g. .doc, .docx, .odt etc.) file. Alternatively you can 'select all' and 'copy' the html output then 'paste' into a new document file.

  13. How to search biocide resistance genes in NGS data using BacMet database?
  14. We have included our in-house developed pipeline/software (BacMet-Scan) for the users who want to investigate biocide- and metal-resistance genes in NGS data sets. This software tool can be downloaded from the Download section. We have provided a tutorial on how to screen the biocide- and metal-resistance genes in the 'tutorials' section.

  15. What are the software requirements for BacMet-Scan?
  16. You will need a Linux/Unix machine with Perl installed. In addition, you will need one or more sequence matching tools. BacMet-Scan supports BLAST, BLAST+, BLAT, pBLAT and vmatch. When those are installed, you should be able to use BacMet-Scan.

  17. How do I install BacMet-Scan?
  18. After installing the required software (see above), Download the BacMet-Scan Perl script and the BacMet database. BacMet-Scan expects to find the database directory in the same directory it is itself located in. Therefore, we recommend to put the script file and the directory and then place a symbolic link to BacMet-Scan in a directory you are allowed to execute software in, e.g. /home/<username>/bin This can be done like this:

    ln -s /home/user/bin/ path_to_directory/BacMet-Scan

    You should then be able to run BacMet-Scan by typing 'BacMet-Scan -h' at the command line.

  19. Can I use BacMet-Scan on Windows?
  20. The easy answer is no. However, if you are willing to give Cygwin, the Windows version of Perl and the Windows version of BLAST a try, we won't stop you. On the other hand, we can't help you much either. We recommend finding a Linux/Unix/Mac OS X machine and use it from there, or run Linux in a virtual machine on your Windows computer (e.g. VirtualBox).

  21. Can I use BacMet-Scan on MacOS X?
  22. Yes. Provided that you have Perl, and one of the sequence matching tools (BLAST, BLAT, vmatch) installed.

  23. BacMet-Scan does not work, and gives an error message beginning with "ERROR! Could not locate BLAST binaries!"
  24. This means that you do not have a working version of BLAST installed (or that BacMet-Scan cannot access it). If you use BacMet-Scan with the default option ('-blast'), it tries to located BLAST+ binaries (blastx or blastp). Try to type 'blastx' or 'blastp' in the terminal and see if you get a "command not found" error message. If so, head to NCBI and download the BLAST+ package.

    If you instead have used the legacy BLAST engine (BacMet-Scan option '-blastall'), you should try to type 'blastall' at the terminal and see if you get the "command not found" error message. If you do, you can download the legacy executables from the same link given above.

  25. I have a really large data set. BacMet-Scan seems to be using BLAST for matching, is any faster tools supported by BacMet-Scan?
  26. Yes! BacMet-Scan supports BLAST, BLAT and Vmatch. Vmatch is good for identical and near-identical matches, where it is super-fast, while BLAT is more useful for sequences that have mismatches and insertions/deletions. Vmatch can be obtained from here, and BLAT can be obtained from here. Both these tools require licensing (free for academic use). BLAT is activated using the BacMet-Scan option '-blat', and vmatch by using the '-vmatch' option.

  27. Both BLAST and BLAT have multi-threaded implementations. Are those supported by BacMet-Scan?
  28. Yes. For BLAST (options '-blast' and '-blastall'), those are activated by specifying the '-cpu' option to something larger than 1 (e.g. '-cpu 4' will utilize four cores). For BLAT, you need to use the '-cpu' option together with the '-pblat' option. You also need to have the parallel BLAT executable installed, which is not maintained by UCSC. PBLAT can be downloaded from here.

  29. I have a set of predicted ORFs, is it possible to scan those with BacMet-Scan?
  30. Yes. Use the '-protein' option to enable comparisons of protein sequences instead of translated nucleic acid sequences.

  31. I want to scan multiple datasets with BacMet-Scan and compare the results. Is there some convenient output option for that?
  32. We have tried to design the software with this mind. Therefore, you can use the '-matrix' option to output a single row of gene counts (one number of each type in the BacMet database). All the output files for different datasets can then be concatenated using the Unix command 'paste' (or by copy-pasting in e.g. MS Excel). Since the genes will be in the same order for all datasets, and zeros will be filled-in for not-found entries, comparison should then be straightforward. To get the gene names in the same order, you can use the '-counts' options for one of the files, which also adds a column with the gene names (in the same order).

  33. I want to see the actual BLAST (or BLAT/vmatch) report. Is that possible?
  34. Yes, that report is saved if you use the -report' option.

  35. There seem to be a lot of different output options; can I get all of them in some convenient way?
  36. The easiest way of doing that is to use the '-all' option, which saves all possible output files.

  37. Does the database give a PubMed reference for each gene?
  38. Yes. Every individual gene in the 'BacMet Experimental database' directly links to references in PubMed. On the other hand, the genes in the 'BacMet Predicted database' are not directly linked to any PubMed references, but all the genes are linked to NCBI protein database.

  39. How to cite BacMet database?
  40. If you found BacMet database useful for your research then you can cite the database as below:

    Pal, C., Bengtsson-Palme, J., Rensing, C., Kristiansson, E., Larsson, DGJ. (2014) BacMet: antibacterial biocide and metal resistance genes database, Nucleic Acids Research, 42, D737-D743. doi: 10.1093/nar/gkt1252

  41. Who should I contact if I have any questions about the database?
  42. You may contact prof Joakim Larsson.




BacMet database/website was developed and designed by Chandan Pal and currently maintained by Joakim Larsson's team

Copyright © 2013-2018 All rights reserved

GU logo