Skip to content

Commit

Permalink
populating projects, small fix for marals people page (#9)
Browse files Browse the repository at this point in the history
* adding two more papers manually for Maral

* adding gecco preprint

* fixing handle for Quinten

* updating marals page

* populating projects

* deleting dummy files
  • Loading branch information
karchern authored Jan 22, 2024
1 parent 448bba7 commit c8fbe98
Show file tree
Hide file tree
Showing 12 changed files with 242 additions and 61 deletions.
40 changes: 40 additions & 0 deletions papers/_posts/2021-05-04-carroll-accurate-de-novo.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,40 @@
---
layout: paper
title: "Accurate de novo identification of biosynthetic gene clusters with GECCO"
nickname: 2021-05-04-carroll-accurate-de-novo
authors: "Carroll LM, Larralde M, Fleck JS, Ponnudurai R, Milanese A, Cappio E, Zeller G"
year: "2021"
journal: "biorxiv"
volume:
issue:
pages:
is_published: true
image: /assets/images/papers/biorxiv.png
projects:
tags: []

# Text
fulltext:
pdf:
pdflink:
pmcid:
preprint:
supplement:

# Links
doi: "10.1101/2021.05.03.442509"
pmid:

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

# Abstract

Biosynthetic gene clusters (BGCs) are enticing targets for (meta)genomic mining efforts, as they may encode novel, specialized metabolites with potential uses in medicine and biotechnology. Here, we describe GECCO (GEne Cluster prediction with COnditional random fields; https://gecco.embl.de), a high-precision, scalable method for identifying novel BGCs in (meta)genomic data using conditional random fields (CRFs). Based on an extensive evaluation of de novo BGC prediction, we found GECCO to be more accurate and over 3x faster than a state-of-the-art deep learning approach. When applied to over 12,000 genomes, GECCO identified nearly twice as many BGCs compared to a rule-based approach, while achieving higher accuracy than other machine learning approaches. Introspection of the GECCO CRF revealed that its predictions rely on protein domains with both known and novel associations to secondary metabolism. The method developed here represents a scalable, interpretable machine learning approach, which can identify BGCs de novo with high precision.
31 changes: 0 additions & 31 deletions projects/_posts/2016-09-25-abcd.md

This file was deleted.

27 changes: 0 additions & 27 deletions projects/_posts/2018-10-05-seaas.md

This file was deleted.

Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Gut dysbiosis / disease associations"
contributors: [qrducarmon]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: How can the microbiome be modulated to improve human health?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

The plastic nature of the gut microbiome facilitates its modulation through non-invasive methods, which makes it an attractive intervention target. Given the gut microbiome’s influence on health and disease, microbiome modulation in principle aims at reverting disease-associated constellations (dysbiosis) back into a healthy state (eubiosis). However, despite a wealth of human microbiome studies, we still do not have a quantitative definition of dysbiosis and eubiosis, not a precise understanding of how diet and lifestyle factors, including medication, shape the gut microbiome. In the past we have collaborated with other groups at EMBL to characterise the effects of antibiotics and non-antibiotic drugs on commensal gut microbes in vitro and are now shifting focus from these chemically defined to more complex dietary interventions and study their effects on the gut microbiome. To model the impact of dietary effects, including fasting, we analyse gut metagenomic and other -omics data in collaborative settings with clinical partners and perform meta-analyses on dietary intervention studies. With this work, we ultimately aim to more rationally design dietary interventions to effectively restore gut health (eubiosis) in an individual-specific way.
27 changes: 27 additions & 0 deletions projects/_posts/2023-11-01-gut-dysbiosis-disease-association.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Gut dysbiosis / disease associations"
contributors: [nkarcher, shitut]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: Can we obtain a quantitative understanding of gut dysbiosis to rationalise microbiome modulation?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

Recent reports suggest that many tumours, in particular those close to the gastrointestinal tract harbour a characteristic microbiota. However, the etiological roles of individual microbes, some of which have been shown to invade host cells, produce genotoxins, or reprogram antitumor immunity, remain poorly understood. We have worked on unravelling the composition of intra-tumoral microbes in several cancer entities through both sequencing and imaging-based methods. We aim to further investigate the cellular and molecular interactions between tumour-resident microbes and host cells and their molecular programs by integrating microbiome-specific assays with host readouts, such as RNA-Seq. Our research was initially focused on colorectal cancer and now increasingly extends to probing the presence and influence of microbes in other tissues, including pancreatic, lung, and liver cancers.
27 changes: 27 additions & 0 deletions projects/_posts/2023-11-01-intratumoral-microbiome.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Intratumoral microbiome"
contributors: [nkarcher, shitut]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: Can we delineate intra-tumoral microbiomes and link tissue microbiome composition and function to cellular and molecular tumour features?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

Recent reports suggest that many tumours, in particular those close to the gastrointestinal tract harbour a characteristic microbiota. However, the etiological roles of individual microbes, some of which have been shown to invade host cells, produce genotoxins, or reprogram antitumor immunity, remain poorly understood. We have worked on unravelling the composition of intra-tumoral microbes in several cancer entities through both sequencing and imaging-based methods. We aim to further investigate the cellular and molecular interactions between tumour-resident microbes and host cells and their molecular programs by integrating microbiome-specific assays with host readouts, such as RNA-Seq. Our research was initially focused on colorectal cancer and now increasingly extends to probing the presence and influence of microbes in other tissues, including pancreatic, lung, and liver cancers.
27 changes: 27 additions & 0 deletions projects/_posts/2023-11-01-microbiome-profiling.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Microbiome profiling"
contributors: [nkarcher, qrducarmon, sromano, fspringer]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: How can we accurately identify and quantify microbes and their gene functions from various sequencing readouts including ones generated from low-biomass samples?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

We are interested in identifying and quantifying microbial taxa and gene functions with high accuracy from various sequencing readouts. Together with the Sunagawa group at ETH Zurich, we have developed and maintained the mOTUs software for high-precision taxonomic profiling from shotgun metagenomic and metatranscriptomic data. Leveraging this expertise, we are now also working on microbial profiling methodologies for low-biomass samples addressing the associated challenges of contamination and low signal-to-noise ratio. Increasingly our focus is shifting towards profiling microbial gene functions and pathways from meta-omics data. We have for instance worked on a software tool (Cayman) for profiling microbial complex carbohydrate metabolism and are in the process of extending our efforts to other relevant areas of secondary metabolism.
31 changes: 31 additions & 0 deletions projects/_posts/2023-11-01-secondary-metabolism.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,31 @@
---
layout: project
title: "Gut dysbiosis / disease associations"
contributors: [lmcarroll]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: How does gut microbial secondary metabolism impact human diseases and drug treatments?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

Microbes of all environments possess an enormous enzymatic repertoire to produce secondary metabolites or chemically modify environmental chemicals. Microbial secondary metabolism has evolved to allow microbes to interact with their environment in complex ways, and the human gut microbiome is no exception.

The production of secondary metabolites, including antibiotics, may give bacteria a competitive advantage in colonising the host. While this can in some cases be advantageous for the host, in other cases, bacteria colonising host tissues are known to produce toxins which can also damage host cells. To characterise the highly diverse secondary metabolism of microbial communities, including that of the gut microbiome, our group develops machine learning methods for the annotation of (meta-)genomic sequences. We apply these methods to discover novel biosynthetic pathways in the human gut, to predict their function and to link some of the resulting natural products with human diseases such as colorectal cancer. Our new method, GECCO, is state of the art for large-scale de novo biosynthetic gene cluster discovery.

Another important aspect of microbial secondary metabolism concerns the biotransformation of xenobiotics. Gut bacterial xenobiotic metabolism has been known to also affect some human medications. In our research we aim to decipher how gut bacteria alter exposure to immunosuppressive drugs in kidney transplant patients. Using both reductionistic experimental studies and extensive clinical trials, our team combines bioinformatics, experimental techniques, and clinical insights. With this multidisciplinary approach we are unravelling the complex drug-bacterial interactions to predict how each patient with their individual-specific gut microbiome responds distinctly to the same medication. Ultimately, we aim to translate these insights to move towards more personalised immunosuppressive therapies with improved efficacy and safety.
27 changes: 27 additions & 0 deletions projects/_posts/2023-11-01-spatial-community-analysis.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Spatial community analysis"
contributors: [nkarcher, shitut]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: How to image diverse microbial communities in their natural environment to reveal spatial community architectures?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

The culture-independent study of microbial communities has been driven by metagenomic sequencing. While this can reveal overall bacterial community composition, all spatial information is generally lost during DNA extraction. However, the spatial organisation of microbial communities is key to their function. Consequently, a better mechanistic understanding of microbe-microbe and microbe-host interactions critically depends on the availability of spatial technologies. In our group we are developing a multiplexed FISH-based assay to detect different bacteria simultaneously in situ. We are currently primarily applying this technology to understand the spatial architecture of bacterial communities in colorectal tumour tissue and how this shapes their interactions with tumour and immune cell populations.
27 changes: 27 additions & 0 deletions projects/_posts/2023-11-01-statistical-modeling-meta-analysis.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,27 @@
---
layout: project
title: "Statistical modelling and Meta analysis"
contributors: [nkarcher, fspringer, ihmgonnet]
handle: spatial
status: collection
type: dataset

# Optional
website:
grant:
grant_url:
image:
tagline:
tags: How to effectively utilise statistical modelling and machine learning to delineate microbiome-disease signatures and identify robust biomarkers for disease diagnosis and prognosis?

# Data and code
github:
neurovault:
openneuro:
figshare:
figshare_names:
osf:
---
{% include JB/setup %}

The human gut microbiome (the complex ecosystem of microbes in the human intestines) is recognized as an important factor influencing human health. In the last decade, a plethora of studies have described associations between the gut microbiome and human diseases, manifesting in the gut, such as colorectal cancer, or in other organ systems, such as cardiovascular or neurodegenerative diseases. These studies have sparked interest in the use of the human microbiome for diagnostic and prognostic purposes in the form of biomarkers identified through statistical modelling and machine learning. However, differences in methodologies and lack of thorough statistical assessments have often generated discrepant results across studies. To address these issues, we develop software frameworks for the identification of associations and machine-learning based microbiome disease signatures and biomarkers. Importantly, we engineer them to handle (observed) confounders and make them applicable in meta-analyses allowing researchers to compare findings across cohorts to obtain more robust findings. We have validated our software pipelines across many microbiome-disease association data sets and implemented easy-to-use interfaces.
10 changes: 8 additions & 2 deletions team/_posts/2019-10-01-baghai-arassi-maral.md
Original file line number Diff line number Diff line change
@@ -1,17 +1,23 @@
---
layout: member
title: Maral Baghai Arassi
position: Clinical Scientist
position: Clinician Scientist
handle: [mbaghai]
science_names: [Baghai Arassi M]
image: baghai-arassi-maral.jpg
image:
alumni: false

# social
email: maral.baghai@embl.de
github: []
osf: []
figshare: []
publons:
orcid: 0000-0002-8957-7767
researchgate: Maral-Baghai-Arassi
scholar: 2V5P3WlohpkC
site: []
twitter: []

---
Maral Baghai Arassi, MD, joined the group as a clinician scientist in October 2019. A graduate of the University of Freiburg Faculty of Medicine, she completed her paediatric residency in November 2023 and is currently a fellow in paediatric nephrology at the University Children’s Hospital in Heidelberg. Her research is dedicated to understanding how the gut microbiome influences kidney transplantation using both reductionistic experimental studies and extensive clinical trials. She focuses on the interactions between immunosuppressive drugs and gut bacteria, aiming to personalise immunosuppressive therapy to reduce post-transplant complications.
2 changes: 1 addition & 1 deletion team/_posts/2021-02-01-ducarmon-quinten.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: member
title: Quinten Ducarmon
position: Postdoctoral Researcher
handle: [qrducarmon]
handle: qrducarmon
nickname: []
science_names: [Ducarmon QR]
image: ducarmon-quinten.jpg
Expand Down

0 comments on commit c8fbe98

Please sign in to comment.