@XLR

Husband. Parent. Dr sc @ETH. Bioinformagician. Father of @PacBio HiFi. Fellow BFX/AI. C++. Geek. Digital Artist. Electronic music. Speaks in memes.

 Germany
armintoepfer.com
Joined November 2008
  • Tweets7.3K
  • Following505
  • Followers1.5K
1.9KMedia
After 12 years at @PacBio, I'm proud to announce that I've been promoted to Fellow - moving back to a full-time individual contributor role to work with the great folks across the company! I'm excited to finally have the space to explore the hard problems and even crazier ideas
11
977
Time to build my own specialized agent from scratch.
40
Why is my timeline filled with people leaving ant?
119
Three. The perfect number of different projects I can steer in parallel.
159
This is basically how I've been operating
NFTmachine 🤖
@boobs_scaryOct 23
Cheat code for achieving remarkable things as an individual is to psychotically compare yourselves to entire teams with infinite resources and beat yourself up mentally until you derive clever enough solutions to reach parity with them.
1
2
193
I'm looking for the greatest models running on 4 DGX Sparks.
2
3
366
Bioinformatics is fast again! 🚀 I might have just set a record: 1 min 46 sec to analyze a 30x @PacBio HiFi human genome (90 GB), from reads to SNVs, indels and structural variants. On a single server with one GPU. That's an estimated 26 cents in the cloud.🧵
3
9
87
4,402
more replies
Preview builds, benchmarks, and the commands behind every number except the SV caller: pblab, our experimental playground at PacBio. github.com/PacificBioscience…

GitHub - PacificBiosciences/pblab: pblab: PacBio HiFi analysis tools

pblab: PacBio HiFi analysis tools. Contribute to PacificBiosciences/pblab development by creating an account on GitHub.

github.com
1
4
327
Some of the optimizations went upstream too. The agents made hifiasm 2.9x and minimap2 1.5x faster on a whole genome, with identical output. Thanks to Haoyu Cheng and @lh3lh3 for building these incredible tools. github.com/chhylp123/hifiasm… github.com/lh3/minimap2/pull…

Speed up hifiasm about 3x with identical output by armintoepfer · Pull Request #956 · chhylp123/h...

This branch makes hifiasm 0.25.0-r726 (ec9a8b2) faster without changing what it computes. Every commit keeps the assembly the same: the GFA files and the overlap bins are byte-identical to upstream...

github.com
1
5
261
Today marks the day we'll start looking differently at bioinformatics.
2
238
Give me Opus 5.5 intelligence on 4x DGX Spark and I'm settled for the rest of my career ;)
2
203
With a minimal CLAUDE. md and guards to validate outputs, I can talk to Opus like to a team of the best SWEs in my field. It will outperform any of us. Still needs plenty of steering that I'm willing to provide.
1
4
396
Max Q! I slowed down the pace to focus on memory usage: - 40 seconds for unaligned CRAM (-10s) - 7 GB RAM for unaligned BAM (-11GB) Comparing against pbmm2 and the last checkpoint. Pushing the frontier of bioinformatics is addictive. 2026Q4 shapes up to be banger!
Armin Töpfer@XLROct 8
Lift-off! Aligning 30x HiFi HG002 in less than 60 seconds on 64 cores. - 128c: 33s/59s un/sorted CRAM - CRAM output reduces output size by 2.5x - Better SNP, INDEL, SV F1s - Lower peak memory It's a PoC for now! @lh3lh3 I'll start porting some of it to mm2 soon.
1
5
461
Developing new software in an established field requires constant output testing. Those gates must be FAST to not slow down the development cycle. DeepVariant 40s, hap.py 3s, Truvari 3.5 minutes? No longer! How about 19 seconds with identical results.
3
4
16
1,194
Me every morning to check what my army of agents produced over night.
1
160
Trying to keep up with open-weight models...
129
Lift-off! Aligning 30x HiFi HG002 in less than 60 seconds on 64 cores. - 128c: 33s/59s un/sorted CRAM - CRAM output reduces output size by 2.5x - Better SNP, INDEL, SV F1s - Lower peak memory It's a PoC for now! @lh3lh3 I'll start porting some of it to mm2 soon.
Armin Töpfer@XLROct 7
How much faster can we map and align accurate long reads? I benchmarked pbmm2 on 30× HiFi HG002 against GRCh38 using 64 CPU cores: ~17 minutes. Can we get below 2 minutes while maintaining SNV, INDEL, and SV calling metrics? mm2 was a jump. Is it time to leap?
3
8
2
39
3,059
And another experiment, re-implemented hap.py in C++ to get it from 16 min to 3 seconds. Identical results. What's next?!
1
1
3
210
How much faster can we map and align accurate long reads? I benchmarked pbmm2 on 30× HiFi HG002 against GRCh38 using 64 CPU cores: ~17 minutes. Can we get below 2 minutes while maintaining SNV, INDEL, and SV calling metrics? mm2 was a jump. Is it time to leap?
2
1
1
10
3,113
Load more