Forum:What Are The Most Influential Bioinformatics/Computational Biology Papers Of 2017?

Question

[dago]

A new year is going to end in few days. Many new exciting discoveries/programs/methods have been published.

In consistency with previous Discussions

• What Are The Most Influential Bioinformatics/Computational Biology Papers Of 2015?
• What Are The Most Influential Bioinformatics/Computational Biology Papers Of 2011?

What do you think were the most influential Bioinformatics/Computational Biology Papers/methods/discovery Of 2017?

Answer 1

[Philipp Bayer]

I think the last few months saw a coming-of-age of genomes as direct graphs and not fasta representations, as in

Genome Graphs (Novak et al) https://www.biorxiv.org/content/early/2017/01/18/101378

Genome Graphs and the Evolution of Genome Inference (Paten et al.) https://www.biorxiv.org/content/early/2017/03/14/101816

Fast and Accurate Genomic Analyses using Genome Graphs (Rakocevic et al.) https://www.biorxiv.org/content/early/2017/09/29/194530

Sequence variation aware genome references and read mapping with the variation graph toolkit (Garrison et al.) https://www.biorxiv.org/content/early/2017/12/15/234856

I'm happy to see this happen, a fasta representation of a genome is just that, a representation that drops a lot of information (ploidy, local rearrangements, heteroplasmy etc.) which in turn influences how you think about the problem at hand. (i.e., you may ignore these problems exist).

Next I'm hoping to see this applied in more complex areas, like gene presence/absence in polyploid organisms with lots of homeologous gene copies.

Answer 2

[andrew.j.skelton73]

Short of alignment free/ pseudo alignment methods (arguably 2016) or ONT producing 1Mb reads - This paper has been one of topic in our group recently if you can get over some of the visualisations, the take home messages are quite informative.

GATK4 is likely to be a huge deal for all of us in 2018, but the maths is all dated for 2017, so I'd argue that's a winner (warning: it's intense), take a look here: Mutect, Pair HMM, and Local assembly

Answer 3

[Hussain Ather]

Not a particular paper, but advances in single-cell sequencing deserves a mention https://www.nature.com/news/single-cell-sequencing-made-simple-1.22233

Answer 4

[arta]

Deep learning (Artificial Intelligence) becoming more and more popular in every filed of science including bioinformatics and computational biology. Cancer classification, identification from tissue images can be done in a more accurate, efficient and faster way than pathologist (no offense !!). There are many publications especially from google DeapMind.

Here is the my nomination Dermatologist-level classification of skin cancer with deep neural networks

[Philipp Bayer]

I think deep learning is delivering great stuff in the area of image recognition but only minimal improvements everywhere else.

The much-hyped DeepVariant paper worked by transforming the problem of SNP-calling into a image-based problem (pictures of piled up reads) and even then only delivered relatively small accuracy improvements at bigger computational and complexity cost..

[Chris Fields]

Yes, the DeepVariant paper might be a good candidate for most over-hyped. I was pretty disappointed in it.

Answer 5

[moldach]

Ever since CRISPR-Cas9 was used to edit human cells in 2013 there has been optimism of treating virtually every disease. In 2018 we will likely see a move into the clinic in Europe and perhaps by 2019 in the U.S. There are already trials in China but researchers have yet to publish any early results from those. Overall we may be a bit too over-optimistic for results from these clinical trials.

The extremely volatile patent-debate between Broad Institute and UC Berkeley (which brings up the whole issue of "do patents hinder scientific progress?" is still on-going but already companies are springing up to commercialize the technology.

I think there were a couple of exciting findings in 2017 that will be pushing the field forward:

1) First successful gene-editing human embryos and correcting mutation for a blood disorder

2) Targeting fusion genes to shrink aggressive tumors in mice

3) New in vivo techniques for epigenetic therapies against human diseases in mouse models (e.g. diabetes)

However, researchers will likely have to consider individual genomes as SNPs and INDELS can result in personalized off-target effects (Lessard et al., 2017).

Answer 6

[edward.messick]

Purely from a microbiome study perspective there were many big publications this year... but I'll put a few

I think that CRIPSR has had a huge year, breaking into the clinical space and actually being used successfully to treat adult patients. I think WGS metagenomics is having a big year too and soon we may be leaving amplicon based methods in the past (despite being cheaper and arguably better, at least for bacteria). The future is in WGS methods. We also seem to be establishing better and more consistent methodologies for analyzing microbiome data, whereas last year it seemed that every lab was doing things their own way (still, of course, its like that - but less so).

Anyway, for the microbiome stuff, we are finding stronger links between signatures in our gut microbiome to diseases (neurological and gut) but also in the metabolism and effects of pharmaceutical medicines. We also are finding links in the development of neurological disorders like Autism among numerous other diseases (obesity, diabetes, etc). Here are a few that were published in Nature this year that I think will have an impact on how we think about our microbiome and its interactions with disease/medicine/our immune system:

(1) Dynamics of the human gut microbiome in inflammatory bowel disease - https://www.nature.com/articles/nmicrobiol20174

(2) Parkinson's disease and Parkinson's disease medications have distinct signatures of the gut microbiome - http://onlinelibrary.wiley.com/doi/10.1002/mds.26942/abstract

(3) Metformin alters the gut microbiome of individuals with treatment-naive type 2 diabetes, contributing to the therapeutic effects of the drug - https://www.nature.com/articles/nm.4345

(4) Metagenomic Shotgun Sequencing and Unbiased Metabolomic Profiling Identify Specific Human Gut Microbiota and Metabolites Associated with Immune Checkpoint Therapy Efficacy in Melanoma Patient - http://www.neoplasia.com/article/S1476-5586%2817%2930238-5/fulltext

(5) A communal catalogue reveals Earth’s multiscale microbial diversity - https://www.nature.com/articles/nature24621

There are too many papers for me to sift through right now and make a choice, but I put some I found interesting/were in high impact journals (namely, Nature).