Tag: Illumina

Published paper: benchmarking resistance gene identification

Since F1000Research uses a somewhat different publication scheme than most journals, I still haven’t understood if this paper is formally published after peer review, but I start to assume it is. There have been very little changes since the last version, so hence I will be lazy and basically repost what I wrote in April when the first version (the “preprint”) was posted online. The paper (1) is the result of a workshop arranged by the JRC in Italy in 2017. It describes various challenges arising from the process of designing a benchmark strategy for bioinformatics pipelines in the identification of antimicrobial resistance genes in next generation sequencing data.

The paper discusses issues about the benchmarking datasets used, testing samples, evaluation criteria for the performance of different tools, and how the benchmarking dataset should be created and distributed. Specially, we address the following questions:

  • How should a benchmark strategy handle the current and expanding universe of NGS platforms?
  • What should be the quality profile (in terms of read length, error rate, etc.) of in silico reference materials?
  • Should different sets of reference materials be produced for each platform? In that case, how to ensure no bias is introduced in the process?
  • Should in silico reference material be composed of the output of real experiments, or simulated read sets? If a combination is used, what is the optimal ratio?
  • How is it possible to ensure that the simulated output has been simulated “correctly”?
  • For real experiment datasets, how to avoid the presence of sensitive information?
  • Regarding the quality metrics in the benchmark datasets (e.g. error rate, read quality), should these values be fixed for all datasets, or fall within specific ranges? How wide can/should these ranges be?
  • How should the benchmark manage the different mechanisms by which bacteria acquire resistance?
  • What is the set of resistance genes/mechanisms that need to be included in the benchmark? How should this set be agreed upon?
  • Should datasets representing different sample types (e.g. isolated clones, environmental samples) be included in the same benchmark?
  • Is a correct representation of different bacterial species (host genomes) important?
  • How can the “true” value of the samples, against which the pipelines will be evaluated, be guaranteed?
  • What is needed to demonstrate that the original sample has been correctly characterised, in case real experiments are used?
  • How should the target performance thresholds (e.g. specificity, sensitivity, accuracy) for the benchmark suite be set?
  • What is the impact of these performance thresholds on the required size of the sample set?
  • How can the benchmark stay relevant when new resistance mechanisms are regularly characterized?
  • How is the continued quality of the benchmark dataset ensured?
  • Who should generate the benchmark resource?
  • How can the benchmark resource be efficiently shared?

Of course, we have not answered all these questions, but I think we have come down to a decent description of the problems, which we see as an important foundation for solving these issues and implementing the benchmarking standard. Some of these issues were tackled in our review paper from last year on using metagenomics to study resistance genes in microbial communities (2). The paper also somewhat connects to the database curation paper we published in 2016 (3), although this time the strategies deal with the testing datasets rather than the actual databases. The paper is the first outcome of the workshop arranged by the JRC on “Next-generation sequencing technologies and antimicrobial resistance” held October 4-5 2017 in Ispra, Italy. You can find the paper here (it’s open access).

On another note, the new paper describing the UNITE database (4) has now got a formal issue assigned to it, as has the paper on tandem repeat barcoding in fungi published in Molecular Ecology Resources last year (5).

References and notes

  1. Angers-Loustau A, Petrillo M, Bengtsson-Palme J, Berendonk T, Blais B, Chan KG, Coque TM, Hammer P, Heß S, Kagkli DM, Krumbiegel C, Lanza VF, Madec J-Y, Naas T, O’Grady J, Paracchini V, Rossen JWA, Ruppé E, Vamathevan J, Venturi V, Van den Eede G: The challenges of designing a benchmark strategy for bioinformatics pipelines in the identification of antimicrobial resistance determinants using next generation sequencing technologies. F1000Research, 7, 459 (2018). doi: 10.12688/f1000research.14509.1
  2. Bengtsson-Palme J, Larsson DGJ, Kristiansson E: Using metagenomics to investigate human and environmental resistomes. Journal of Antimicrobial Chemotherapy, 72, 2690–2703 (2017). doi: 10.1093/jac/dkx199
  3. Bengtsson-Palme J, Boulund F, Edström R, Feizi A, Johnning A, Jonsson VA, Karlsson FH, Pal C, Pereira MB, Rehammar A, Sánchez J, Sanli K, Thorell K: Strategies to improve usability and preserve accuracy in biological sequence databases. Proteomics, 16, 18, 2454–2460 (2016). doi: 10.1002/pmic.201600034
  4. Nilsson RH, Larsson K-H, Taylor AFS, Bengtsson-Palme J, Jeppesen TS, Schigel D, Kennedy P, Picard K, Glöckner FO, Tedersoo L, Saar I, Kõljalg U, Abarenkov K: The UNITE database for molecular identification of fungi: handling dark taxa and parallel taxonomic classifications. Nucleic Acids Research, 47, D1, D259–D264 (2019). doi: 10.1093/nar/gky1022
  5. Wurzbacher C, Larsson E, Bengtsson-Palme J, Van den Wyngaert S, Svantesson S, Kristiansson E, Kagami M, Nilsson RH: Introducing ribosomal tandem repeat barcoding for fungi. Molecular Ecology Resources, 19, 1, 118–127 (2019). doi: 10.1111/1755-0998.12944

New preprint: benchmarking resistance gene identification

This weekend, F1000Research put online the non-peer-reviewed version of the paper resulting from a workshop arranged by the JRC in Italy last year (1). (I will refer to this as a preprint, but at F1000Research the line is quite blurry between preprint and published paper.) The paper describes various challenges arising from the process of designing a benchmark strategy for bioinformatics pipelines (2) in the identification of antimicrobial resistance genes in next generation sequencing data.

The paper discusses issues about the benchmarking datasets used, testing samples, evaluation criteria for the performance of different tools, and how the benchmarking dataset should be created and distributed. Specially, we address the following questions:

  • How should a benchmark strategy handle the current and expanding universe of NGS platforms?
  • What should be the quality profile (in terms of read length, error rate, etc.) of in silico reference materials?
  • Should different sets of reference materials be produced for each platform? In that case, how to ensure no bias is introduced in the process?
  • Should in silico reference material be composed of the output of real experiments, or simulated read sets? If a combination is used, what is the optimal ratio?
  • How is it possible to ensure that the simulated output has been simulated “correctly”?
  • For real experiment datasets, how to avoid the presence of sensitive information?
  • Regarding the quality metrics in the benchmark datasets (e.g. error rate, read quality), should these values be fixed for all datasets, or fall within specific ranges? How wide can/should these ranges be?
  • How should the benchmark manage the different mechanisms by which bacteria acquire resistance?
  • What is the set of resistance genes/mechanisms that need to be included in the benchmark? How should this set be agreed upon?
  • Should datasets representing different sample types (e.g. isolated clones, environmental samples) be included in the same benchmark?
  • Is a correct representation of different bacterial species (host genomes) important?
  • How can the “true” value of the samples, against which the pipelines will be evaluated, be guaranteed?
  • What is needed to demonstrate that the original sample has been correctly characterised, in case real experiments are used?
  • How should the target performance thresholds (e.g. specificity, sensitivity, accuracy) for the benchmark suite be set?
  • What is the impact of these performance thresholds on the required size of the sample set?
  • How can the benchmark stay relevant when new resistance mechanisms are regularly characterized?
  • How is the continued quality of the benchmark dataset ensured?
  • Who should generate the benchmark resource?
  • How can the benchmark resource be efficiently shared?

Of course, we have not answered all these questions, but I think we have come down to a decent description of the problems, which we see as an important foundation for solving these issues and implementing the benchmarking standard. Some of these issues were tackled in our review paper from last year on using metagenomics to study resistance genes in microbial communities (3). The paper also somewhat connects to the database curation paper we published in 2016 (4), although this time the strategies deal with the testing datasets rather than the actual databases. The paper is the first outcome of the workshop arranged by the JRC on “Next-generation sequencing technologies and antimicrobial resistance” held October 4-5 last year in Ispra, Italy. You can find the paper here (it’s open access).

References and notes

  1. Angers-Loustau A, Petrillo M, Bengtsson-Palme J, Berendonk T, Blais B, Chan KG, Coque TM, Hammer P, Heß S, Kagkli DM, Krumbiegel C, Lanza VF, Madec J-Y, Naas T, O’Grady J, Paracchini V, Rossen JWA, Ruppé E, Vamathevan J, Venturi V, Van den Eede G: The challenges of designing a benchmark strategy for bioinformatics pipelines in the identification of antimicrobial resistance determinants using next generation sequencing technologies. F1000Research, 7, 459 (2018). doi: 10.12688/f1000research.14509.1
  2. You may remember that I hate the term “pipeline” for bioinformatics protocols. I would have preferred if it was called workflows or similar, but the term “pipeline” has taken hold and I guess this is a battle where I have essentially lost. The bioinformatics workflows will be known as pipelines, for better and worse.
  3. Bengtsson-Palme J, Larsson DGJ, Kristiansson E: Using metagenomics to investigate human and environmental resistomes. Journal of Antimicrobial Chemotherapy, 72, 2690–2703 (2017). doi: 10.1093/jac/dkx199
  4. Bengtsson-Palme J, Boulund F, Edström R, Feizi A, Johnning A, Jonsson VA, Karlsson FH, Pal C, Pereira MB, Rehammar A, Sánchez J, Sanli K, Thorell K: Strategies to improve usability and preserve accuracy in biological sequence databases. Proteomics, 16, 18, 2454–2460 (2016). doi: 10.1002/pmic.201600034

Reflections from the NGS Congress 2016

As the 8th Next Generation Sequencing Congress in London is drawing to a close as I write this, I have a few reflections that might warrant sharing. The first thing that has been apparent this year compared to the two previous times I have visited the event (in 2012 and 2013) is that there was very little talk about where Illumina sequencing is heading next. Instead the discussion was about the applications of Illumina sequencing in the clinical setting; so apparently this is now so mainstream that we only expect slow progress towards longer reads. Apart from that, Illumina is a completed, mature technology. Instead, the flashlight is now pointing entirely towards long-read sequencing (PacBio, NanoPore) as the next big thing. However, the excitement around these technologies has also sort of faded compared to in 2013 when they were soon-to-arrive. Indeed, it seems like there’s not much to be excited about in the sequencing field at the moment, or at least Oxford Global (who are hosting the conference) has failed to get these technologies here.

What also strikes me is the vast amounts of talk about RNAseq of cancer cells. The scope of this event has narrowed dramatically in the past three years. Which makes me substantially less interested in returning next year. If there is not much to be excited about, and the focus is only on cancer sequencing – despite the human microbiota being a very hot topic at the moment – what is the reason for non-cancer researchers to come to the event? There will need to be a stark shift towards another direction of this event if the arrangers want it to remain a broad NGS event. Otherwise, they may just as well go all in and rename the event the Next Generation Sequencing of Cancer Congress. But I hope they choose to widen the scope again; conferences discussing technology as a foundation for a variety of applications are important meeting points and spawning grounds for novel ideas.

Published paper: The global resistome

Late yesterday, Microbiome put online our most recent work, covering the resistomes to antibiotics, biocides and metals across a vast range of environments. In the paper (1), we perform the largest characterization of resistance genes, mobile genetic elements (MGEs) and bacterial taxonomic compositions to date, covering 864 different metagenomes from humans (350), animals (145) and external environments such as soil, water, sewage, and air (369 in total). All the investigated metagenomes were sequenced to at least 10 million reads each, using Illumina technology, making the results more comparable across environments than in previous studies (2-4).

We found that the environment types had clear differences both in terms of resistance profiles and bacterial community composition. Humans and animals hosted microbial communities with limited taxonomic diversity as well as low abundance and diversity of biocide/metal resistance genes and MGEs. On the contrary, the abundance of ARGs was relatively high in humans and animals. External environments, on the other hand, showed high taxonomic diversity and high diversity of biocide/metal resistance genes and MGEs. Water, sediment and soil generally carried low relative abundance and few varieties of known ARGs, whereas wastewater and sludge were on par with the human gut. The environments with the largest relative abundance and diversity of ARGs, including genes encoding resistance to last resort antibiotics, were those subjected to industrial antibiotic pollution and air samples from a Beijing smog event.

A paper investigating this vast amount of data is of course hard to describe in a blog post, so I strongly suggest the interested reader to head over to Microbiome’s page and read the full paper (1). However, here’s a ver short summary of the findings:

  • The median relative abundance of ARGs across all environments was 0.035 copies per bacterial 16S rRNA
  • Antibiotic-polluted environments have (by far) the highest abundances of ARGs
  • Urban air samples carried high abundance and diversity of ARGs
  • Human microbiota has high abundance and diversity of known ARGs, but low taxonomic diversity compared to the external environment
  • The human and animal resistomes are dominated by tetracycline resistance genes
  • Over half of the ARGs were only detected in external environments, while 20.5 % were found in human, animal and at least one of the external environments
  • Biocide and metal resistance genes are more common in external environments than in the human microbiota
  • Human microbiota carries low abundance and richness of MGEs compared to most external environments

Importantly, less than 1.5 % of all detected ARGs were found exclusively in the human microbiome. At the same time, 57.5 % of the known ARGs were only detected in metagenomes from environmental samples, despite that the majority of the investigated ARGs were initially encountered in pathogens. Still, our analysis suggests that most of these genes are relatively rare in the human microbiota. Environmental samples generally contained a wider distribution of resistance genes to a more diverse set of antibiotics classes. For example, the relative abundance of beta-lactam resistance genes was much larger in external environments than in human and animal microbiomes. This suggests that the external environment harbours many more varieties of resistance genes than the ones currently known from the clinic. Indeed, functional metagenomics has resulted in the discovery of many novel ARGs in external environments (e.g. 5). This all fits well with an overall much higher taxonomic diversity of environmental microbial communities. In terms of consequences associated with the potential transfer of ARGs to human pathogens, we argue that unknown resistance genes are of greater concern than those already known to circulate among human-associated bacteria (6).

This study describes the potential for many external environments, including those subjected to pharmaceutical pollution, air and wastewater/sludge, to serve as hotspots for resistance development and/or transmission of ARGs. In addition, our results indicate that these environments may play important roles in the mobilization of yet unknown ARGs and their further transmission to human pathogens. To provide guidance for risk-reducing actions we – based on this study – suggest strict regulatory measures of waste discharges from pharmaceutical industries and encourage more attention to air in the transmission of antibiotic resistance (1).

References

  1. Pal C, Bengtsson-Palme J, Kristiansson E, Larsson DGJ: The structure and diversity of human, animal and environmental resistomes. Microbiome, 4, 54 (2016). doi: 10.1186/s40168-016-0199-5
  2. Durso LM, Miller DN, Wienhold BJ. Distribution and quantification of antibiotic resistant genes and bacteria across agricultural and non-agricultural metagenomes. PLoS One. 2012;7:e48325.
  3. Nesme J, Delmont TO, Monier J, Vogel TM. Large-scale metagenomic-based study of antibiotic resistance in the environment. Curr Biol. 2014;24:1096–100.
  4. Fitzpatrick D, Walsh F. Antibiotic resistance genes across a wide variety of metagenomes. FEMS Microbiol Ecol. 2016. doi:10.1093/femsec/fiv168.
  5. Allen HK, Moe LA, Rodbumrer J, Gaarder A, Handelsman J. Functional metagenomics reveals diverse β-lactamases in a remote Alaskan soil. ISME J. 2009;3:243–51.
  6. Bengtsson-Palme J, Larsson DGJ: Antibiotic resistance genes in the environment: prioritizing risks. Nature Reviews Microbiology, 13, 369 (2015). doi: 10.1038/nrmicro3399-c1

Talk at Swedish Bioinformatics Workshop

I have had the pleasure to be chosen as a speaker for next week’s (ten days from now) Swedish Bioinformatics Workshop. My talk is entitled “Turn up the signal – wipe out the noise: Gaining insights into bacterial community functions using metagenomic data“, and will largely deal with the same questions as my talk on EDAR3 in May this year. As then, the talk will highlight the some particular pitfalls related to interpretation of data, and exemplify how flawed analysis practices can result in misleading conclusions regarding community function, and use examples from our studies of environments subjected to pharmaceutical pollution in India, the effect of travel on the human resistome, and modern municipal wastewater treatment processes.

The talk will take place on Thursday, September 24, 2015 at 16:30. The full program for the conference can be found here. And also, if you want a sneak peak of the talk, you can drop by on Friday 13.00 at Chemistry and Molecular Biology, where I will give a seminar on the same topic in the Monthly Bioinformatic Practical Meetings series.

Talk on the EDAR2015 conference

I will be giving a talk at the Third International symposium on the environmental dimension of antibiotic resistance (EDAR2015) next month (five weeks from now. The talk is entitled “Turn up the signal – wipe out the noise: Gaining insights into antibiotic resistance of bacterial communities using metagenomic data“, and will deal with handling of metagenomic data in antibiotic resistance gene research. The talk will highlight the some particular pitfalls related to interpretation of data, and exemplify how flawed analysis practices can result in misleading conclusions regarding antibiotic resistance risks. I will particularly address how taxonomic composition influences the frequencies of resistance genes, the importance of knowledge of the functions of the genes in the databases used, and how normalization strategies influence the results. Furthermore, we will show how the context of resistance genes can allow inference of their potential to spread to human pathogens from environmental or commensal bacteria. All these aspects will be exemplified by data from our studies of environments subjected to pharmaceutical pollution in India, the effect of travel on the human resistome, and modern municipal wastewater treatment processes.

The talk will take place on Monday, May 18, 2015 at 13:20. The full scientific program for the conference can be found here. Registration for the conference is still possible, although not for the early-bird price. I look forward to see a lot of the people who will attend the conference, and hopefully also you!

Published paper: Antibiotic resistance genes in a polluted lake

The first work in which I have employed metagenomics to investigate antibiotic resistance has been accepted in Frontiers in Microbiology, and is (at the time of writing) available as a provisional PDF. In the paper (1), which is co-authored by Fredrik Boulund, Jerker Fick, Erik Kristiansson and Joakim Larsson, we have used shotgun metagenomic sequencing of an Indian lake polluted by dumping of waste from pharmaceutical production. We used this data to describe the diversity of antibiotic resistance genes and the genetic context of those, to try to predict their genetic transferability. We found resistance genes against essentially every major class of antibiotics, as well as large abundances of genes responsible for mobilization of genetic material. Resistance genes were estimated to be 7000 times more abundant in the polluted lake than in a Swedish lake included for comparison, where only eight resistance genes were found. The abundances of resistance genes have previously only been matched by river sediment subject to pollution from pharmaceutical production (2). In addition, we describe twenty-six known and twenty-one putative novel plasmids from the Indian lake metagenome, indicating that there is a large potential for horizontal gene transfer through conjugation. Based on the wide range and high abundance of known resistance factors detected, we believe that it is plausible that novel resistance genes are also present in the lake. We conclude that environments polluted with waste from antibiotic manufacturing could be important reservoirs for mobile antibiotic resistance genes. This work further highlights previous findings that pharmaceutical production settings could provide sufficient selection pressure from antibiotics (3) to drive the development of multi-resistant bacteria (4,5), resistance which may ultimately end up in pathogenic species (6,7). The paper can be read in its entirety here.

References:

  1. Bengtsson-Palme J, Boulund F, Fick J, Kristiansson E, Larsson DGJ: Shotgun metagenomics reveals a wide array of antibiotic resistance genes and mobile elements in a polluted lake in India. Frontiers in Microbiology, Volume 5, Issue 648 (2014). doi: 10.3389/fmicb.2014.00648
  2. Kristiansson E, Fick J, Janzon A, Grabic R, Rutgersson C, Weijdegård B, Söderström H, Larsson DGJ: Pyrosequencing of antibiotic-contaminated river sediments reveals high levels of resistance and gene transfer elements. PLoS ONE, Volume 6, e17038 (2011). doi:10.1371/journal.pone.0017038.
  3. Larsson DGJ, de Pedro C, Paxeus N: Effluent from drug manufactures contains extremely high levels of pharmaceuticals. J Hazard Mater, Volume 148, 751–755 (2007). doi:10.1016/j.jhazmat.2007.07.008
  4. Marathe NP, Regina VR, Walujkar SA, Charan SS, Moore ERB, Larsson DGJ, Shouche YS: A Treatment Plant Receiving Waste Water from Multiple Bulk Drug Manufacturers Is a Reservoir for Highly Multi-Drug Resistant Integron-Bearing Bacteria. PLoS ONE, Volume 8, e77310 (2013). doi:10.1371/journal.pone.0077310
  5. Johnning A, Moore ERB, Svensson-Stadler L, Shouche YS, Larsson DGJ, Kristiansson E: Acquired genetic mechanisms of a multiresistant bacterium isolated from a treatment plant receiving wastewater from antibiotic production. Appl Environ Microbiol, Volume 79, 7256–7263 (2013). doi:10.1128/AEM.02141-13
  6. Pruden A, Larsson DGJ, Amézquita A, Collignon P, Brandt KK, Graham DW, Lazorchak JM, Suzuki S, Silley P, Snape JR., et al.: Management options for reducing the release of antibiotics and antibiotic resistance genes to the environment. Environ Health Perspect, Volume 121, 878–885 (2013). doi:10.1289/ehp.1206446
  7. Finley RL, Collignon P, Larsson DGJ, McEwen SA, Li X-Z, Gaze WH, Reid-Smith R, Timinouni M, Graham DW, Topp E: The scourge of antibiotic resistance: the important role of the environment. Clin Infect Dis, Volume 57, 704–710 (2013). doi:10.1093/cid/cit355

PetKit updated to version 1.1

It’s been a while since the PETKit got any attention from me. Partially, that has been due to a nasty bug that could produce no output for one of the read files in Pefcon when using FASTA input files, but mostly it has simply been due to lack of time to continue development on the package. Now, I have finally put all threads together (bug fixes, new features, documentation) and today the 1.1 version is released! The new features are:

  • A new tool has been added – peacat – that can be used to e.g. stitch contigs together that have been separated for one reason or another in an assembly
  • Another tool – pemap – has been added that can be used to determine whether an assembled contig is from a circular DNA element
  • The default offset value for FASTQ files has been set to 33 (as in Sanger and Illumina 1.8+ PHRED format)
  • The documentation has been vastly improved (but is still rather inferior)

PETKit updated – Critical bug fix

Some good and some bad news regarding the PETKit. Good news first; I have written a fourth tool for the PETKit, which is included in the latest release (version 1.0.2b, download here). The new tool is called Pesort, and sorts input read pairs (or single reads) so that the read pairs occur in the same order. It also sorts out which reads that don’t have a pair and outputs them to a separate file. All this is useful if you for some reason have ended up with a scrambled read file (pair). This can e.g. happen if you want to further process the reads after running Khmer or investigate the reads remaining after mapping to a genome.

Then the bad news. There’s a critical bug in PETKit version 1.0.1b. This bug manifest itself when using custom offsets for quality scores (using the –offset option), and makes the Pearf and Pepp tools too strict – leading to that they discard reads that actually are of good quality. This does not affect the Pefcon program. If you use the PETKit for read filtering or ORF prediction, and have used custom offset values, I recommend that you re-run your data with the newly released PETKit version (1.0.2b), in which this bug has been fixed. If you have only used the default offset setting, your safe. I sincerely apologize for any inconveniences that this might have caused.