Notes on mouse ibd

[MatthewRalston]

Reviewing a irritant-on-mouse NOD lineage study on Altered Schaefler formula/flora.

[MatthewRalston]

This is taking a back seat to my productivity stack, org mode, Emacs, journaling/ELN setup, blog posts on other civic topics, and a blog post on an anemic dataset from M. domestica on PFOA/PFOS forever chemical treatment on whole blood. Comparing aligners used, expression matrices, and DEGs.

More to come. But this case really hasn't progressed much after doing alignment. The program does what I want it to do. I need more time and mentorship to flesh out the manuscript and population comparisons.

I feel....discarded.

[MatthewRalston]

Findings

To summarize on the sub issue development today, I decided to briefly mention that the original concept here got derailed by readings and practicals surrounding genome assembly algorithms the alignment topic, the concept of microflora populations, and the concept of assembling equal proportions of close to 11-18 different species by genomic configs, equal read proportions generated by a classical read simulator, and the use of some test sequences from the microbial species selections to test alignment and assembly procedures that may be developed at all

NW alignment

development surrounding two subroutines unfinished yet topical.

** "Alignment" **

a classical NW seed extend sub problem with scoring semantics, over top of a

** "minimizers"**

routine to make a kmer subset for binning alignment threading if needed (i.e. top-level of sub-problem categorizations, recursive depth, and separation of concerns, heuristics, scoring, and sub-space bra ch-and-bound on problem size growth as each dataset would be assessed, and uniform or "normal", in today's ML parlance, treatment of equal needs in the problem

and for minmax on the alignment/assembly category of su problems the routine would like to utilize for a seed-explore, then branch and extend as a sub problem as the solution trivializes via knowledge of needs by disparate test datasets.

Thus, a NW alignment approach, not an alignment algorithm that solves, by any guarantee or warrantee of any kind, the "local", global, or feasible solution a single match.

Incomplete on mins

The minimizers is incomplete because it involves no sequence based heuristics for disparate, adaptive, or progressive minimzer selection rate or taper.

The minimizers is incomplete because either a global, uniform* (the current approach), or useful format describing the selection has been proposed. It's a redundant, useless extension of a tabular minimal tuple describing the single base, coordinate, and minimal metadata for the "step" at hand in trivial selection operations on kmer identities from .kdb files.

disclaimer on possible extensions

Unclear if linear algebra libraries need to be utilized for specialized compute needs.

Memory is a huge issue in 2026.