Sign in

Igor Martayan

@imartayan.bsky.social
838 followers 361 following 97 posts

PhD in algorithmic bioinformatics at @bonsaiseqbioinfo.bsky.social. Interested in space-efficient data structures, sketching algorithms & high-performance computing igor.martayan.org

PostsRepliesMedia
Igor Martayan @imartayan.bsky.social · 27/04/2025
The last speaker is Liang Hong on the detection of protein homologs using deep dense retrieval. www.nature.com/articles/s41...
000 0
Igor Martayan @imartayan.bsky.social · 27/04/2025
Time for the highlight talks starting with Kunhyung Bahk on SigAlign, an alignment algorithm guided by explicit similarity criteria. academic.oup.com/nar/article/...
100 1
Igor Martayan @imartayan.bsky.social · 27/04/2025
Now Adam Park from @david-koslicki.bsky.social's group introduces Prokrustean graphs, a new approach for size-agnostic k-mer analysis. www.biorxiv.org/content/10.1...
100 1
Igor Martayan @imartayan.bsky.social · 27/04/2025
Ghanshyam Chandra is now presenting a new alignment-free genotyping framework for pangenomes. www.biorxiv.org/content/10.1...
100 1
Igor Martayan @imartayan.bsky.social · 27/04/2025
Next up is Nathaniel Brown from @benlangmead.bsky.social's group presenting col-bwt, a new algorithm for computing chain statistics using multi-maximal unique matches. www.biorxiv.org/content/10.1...
2165 16
Igor Martayan @imartayan.bsky.social · 27/04/2025
The sequence analysis session of #RECOMB2025 is off to a great start with @jimshaw.bsky.social presenting devider, a new algorithm for haplotyping small sequences from long-read sequencing. www.biorxiv.org/content/10.1...
1266 26
Igor Martayan @imartayan.bsky.social · 26/04/2025
I'll be presenting our work on hyper-k-mers at #RECOMB today at 10:40 KST! You can get a sneak peek at the slides here: igor.martayan.org/slides-recom... Come say hi if you'd like to chat, or just get one of these cute stickers!
A pack of stickers with the logo of our new tool, K-mer Fast Counter a.k.a. KFC
1184 18
Igor Martayan @imartayan.bsky.social · 25/04/2025
@camillemrcht.bsky.social is back on stage to present Vizitig, an interactive visualization tool for sequencing datasets developped by @bdegardins.bsky.social
146 6
Igor Martayan @imartayan.bsky.social · 25/04/2025
This time, @camillemrcht.bsky.social is on the other side of the stage to present the next version of REINDEER, a scalable index for k-mer abundances
131 3
Igor Martayan @imartayan.bsky.social · 29/01/2025
The main idea of our approach is to stream multiple chunks of the sequence to process them in parallel, coupled with a nearly branchless algorithm to compute sliding window minima, fully implemented using SIMD instructions.
Overview of the vectorized computation of minimizers. The sequence is first split into L chunks that are processed in parallel. The bases of each chunk are gathered in a SIMD register and passed to a vectorized rolling hash function that computes a hash for the k-mers in each lane. The position of the minimum is then computed over a sliding window of w vectors of hashes. The positions are finally reordered and deduplicated to have a unique occurrence of each position.
140 4