Update README.md

README.md

@@ -117,41 +117,6 @@ Example output:
 
 Result #3 - Population Informative TR Loci
 ==========================================
-The overall PCA of the hprc_105 database shows:
-
-![alt text](https://github.com/collaborativebioinformatics/tandemrepeats/blob/main/imgs/baseline_pca_tr_105samples.png?raw=true)
-
-We explored if it is possible to find a subset of TR loci that could generate an equivalent amount of separation by
-population in a PCA.
 
 See [this notebook](https://github.com/collaborativebioinformatics/tandemrepeats/blob/main/English_EDA/MainNotebook.ipynb) for details.
 
-# Old README
-# Outlines ![alt text](https://github.com/collaborativebioinformatics/tandemrepeats/blob/main/imgs/Slide1.png?raw=true)
-![alt text](https://github.com/collaborativebioinformatics/tandemrepeats/blob/main/imgs/Slide2.png?raw=true)
-
-Background
-===========
-For a tutorial on using tdb for programmatic access, see Introduction notebook. The motivation for tdb was that TRs can be better represented as ‘replacements’ of reference sequence spans with contracted/expanded alternate allele sequence. This type of representation removes alignment ambiguities, which TRs are highly susceptible to. Furthermore, VCFs are not a normalized data structure. Each ‘row’ in a VCF can hold multiple alleles and multiple samples. This, combined with the mixed data-types, makes parsing VCF files… unpleasant.  The tdb is a normalized database with three tables with information on loci, alleles, and samples. The data can be parsed by standard data science libraries, such as pandas, with ease.
-
-Methods
-============
-
-Query #1 - Population Structure
---------------
-There is already a population structure notebook which will identify loci with >= 20 alleles and plot a clustermap of how similar samples’ alleles are. This comes with a clustering that - at least in the hprc example data - reconstructs the population structure pretty well. We can refine this query by:
-Improving the selection of loci: Is just >=20 alleles sufficient or could we leverage length/sequence polymorphism queries to get a more informative set of loci?
-Improving reporting of population structure: The clustermap is cool, but that’s not parsable. Writing a “Sample->ClusterID”, could be more informative.
-Could this query be expanded to perform a TR-specific kinship analysis similar to plink’s kinship? Could a subset of TRs be as powerful (more powerful??) as genome-wide SNPs?
-
-Query #2 - PCA
---------------
-Similar to the population structure, there is already an example notebook which will perform a PCA on a tdb. This could be improved. This could also be expanded to perform PCA on methylation data. There may be population structures to dna methylation data. If we can show that they line up to dna variants’ population structure would be neat.
-
-Query #3 - Length Outliers
---------------
-Again, there is an example notebook for finding TR alleles which have an anomalous length. There are other approaches which work to find length outliers. A comprehensive single report of all of these measures would assist researchers in prioritizing tandem repeats.
-
-Query #4 - TR Structure
---------------
-Given the multiple TR alleles over a locus, we can annotate the TR motifs on each sequence and perform an MSA. We can then consolidate and create a ‘consensus’ structure of the repeats over the spans. This output should allow more detailed analysis of length outliers because we would no longer be just looking at the length of sequence over the locus but have motifs and copy numbers aligned across alleles. A light-weight notebook that leverages abpoa and tr-solve to build some of this information is already available. However, we’d want to replace tr-solve for annotating motifs. TRF is possible, but it will redundantly annotate spans which would make deconvolution of the repeat structure over multiple sequences difficult.