DiegoFE
@diegofe94.bsky.social
| Bioinformatician | Data scientist | PhD Student at MALL Lab (UDC) |
To align platforms, we tested Batch Effect Removal (BER) methods.
✅ Good: Limma & FAbatch reduce platform noise while preserving disease signal
⚠️ Caution: Some (e.g., Combat w/ ref) may overfit & distort biology
No universal method yet—but promising paths ahead!
#Microbiome #Bioinformatics 👇
✅ Good: Limma & FAbatch reduce platform noise while preserving disease signal
⚠️ Caution: Some (e.g., Combat w/ ref) may overfit & distort biology
No universal method yet—but promising paths ahead!
#Microbiome #Bioinformatics 👇
May 19, 2025 at 8:49 AM
To align platforms, we tested Batch Effect Removal (BER) methods.
✅ Good: Limma & FAbatch reduce platform noise while preserving disease signal
⚠️ Caution: Some (e.g., Combat w/ ref) may overfit & distort biology
No universal method yet—but promising paths ahead!
#Microbiome #Bioinformatics 👇
✅ Good: Limma & FAbatch reduce platform noise while preserving disease signal
⚠️ Caution: Some (e.g., Combat w/ ref) may overfit & distort biology
No universal method yet—but promising paths ahead!
#Microbiome #Bioinformatics 👇
🧬 We benchmarked taxonomic merging strategies using a colorectal cancer dataset (740 patients).
Using GreenGenes2, a unified DB, we saw:
• 2× more genera
• 3× more species
• Higher taxa retention than other approaches
GG2 boosts overlap between 16S & WGS data.
📊👇
Using GreenGenes2, a unified DB, we saw:
• 2× more genera
• 3× more species
• Higher taxa retention than other approaches
GG2 boosts overlap between 16S & WGS data.
📊👇
May 19, 2025 at 8:49 AM
🧬 We benchmarked taxonomic merging strategies using a colorectal cancer dataset (740 patients).
Using GreenGenes2, a unified DB, we saw:
• 2× more genera
• 3× more species
• Higher taxa retention than other approaches
GG2 boosts overlap between 16S & WGS data.
📊👇
Using GreenGenes2, a unified DB, we saw:
• 2× more genera
• 3× more species
• Higher taxa retention than other approaches
GG2 boosts overlap between 16S & WGS data.
📊👇
We´re happy to share a new conference paper from the #MALL lab!
We explore integrating microbiome data from 16S rRNA & Whole Genome Shotgun across platforms. Can we combine them meaningfully despite their differences?
Lecture Notes in Computer Science from #CIBB2024.
📖 doi.org/10.1007/978-...
🧵👇
We explore integrating microbiome data from 16S rRNA & Whole Genome Shotgun across platforms. Can we combine them meaningfully despite their differences?
Lecture Notes in Computer Science from #CIBB2024.
📖 doi.org/10.1007/978-...
🧵👇
May 19, 2025 at 8:49 AM
We´re happy to share a new conference paper from the #MALL lab!
We explore integrating microbiome data from 16S rRNA & Whole Genome Shotgun across platforms. Can we combine them meaningfully despite their differences?
Lecture Notes in Computer Science from #CIBB2024.
📖 doi.org/10.1007/978-...
🧵👇
We explore integrating microbiome data from 16S rRNA & Whole Genome Shotgun across platforms. Can we combine them meaningfully despite their differences?
Lecture Notes in Computer Science from #CIBB2024.
📖 doi.org/10.1007/978-...
🧵👇