Bill Pearson

[Bill Pearson @FASTABill]

@aaronquinlan @UUtah Congratulations. UUtah is lucky to have you.

[Bill Pearson @FASTABill]

@deniswirtz @mike_schatz It should be pointed out that those “Institutional Funds” are largely derived from indirect costs from Federal grants.

[Bill Pearson @FASTABill]

@bioinformer @StevenSalzberg1 @crill_labs Genbank, since its inception, has the policy of being and "archival" database that contains what was submitted to it. "Contaminants" in Genbank have been well documented for more than 30 years. That is why there are also "curated" databases at the NCBI, like RefSeq.

[Bill Pearson @FASTABill]

@StevenSalzberg1 @JillFilipovic @lizditz Or maybe Israel's killing 10's of thousands of Palestinians in Gaza is giving Hamas exactly what they want. It will certainly keep hate alive for another generation or two. Hard to know.

[Bill Pearson @FASTABill]

@StevenSalzberg1 Actually, the paper is https://t.co/KaboNDf7R0
But it’s nice to be (mis)-remembered.

[Bill Pearson @FASTABill]

@phylogenomics (4) Don’t bother looking for ORFs - BLASTX and FASTX do that for you.
(5) Make sure you are doing DNA reads vs protein databases (BLASTX, FASTX) - 5 - 10x more sensitive.

[Bill Pearson @FASTABill]

@phylogenomics (2) Search with shallower scoring matrices. PAM70 for BLASTX and VT80 for FASTX. These work much better for shorter queries.
(3) Don’t search pFam or NR/Refseq. It is much more sensitive to search smaller databases that are targeted

[Bill Pearson @FASTABill]

@phylogenomics You do not need anything except computer time. Use BLASTX or FASTX to compare your reads (as DNA) against a phylogenetically appropriate protein sequence database. For bacteria, pick 20 - 50 well annotated bacterial proteomes (1-2 million proteins). /1

[Bill Pearson @FASTABill]

@CWolberger I can imagine the problem might be that some systems (or less technically astute reviewers) may automatically follow the links, thus potentially de-anonymizing the review process. Particularly one can put more information in the link than is displayed in the text.

[Bill Pearson @FASTABill]

What does ChatGPT know about the difference between FASTA and BLAST? It does an excellent job, as long as your substitute BLAST for FASTA (and vice versa) for most of the result. Both are heuristic, FASTA is slower, more accurate, and offers global alignment.

[Stephanie Hicks, PhD @stephaniehicks]

@LuciaScience @mikelove @daniela_witten @drisso1893 A few years back, @pkkimes @keegankorthauer, myself, and others wrote a review / benchmark of various flavors of FDR in application to comp bio problems. Let me know if this helps, or I can try to summarize here too!

https://t.co/US66Ww3PiV

[Bill Pearson @FASTABill]

@loyalgoff BlastP with the protein encoded by that mRNA against the proteins in the other species, take the significant hits, and get their mRNAs.

[Geoff Barton @gjbarton]

Margaret Dayhoff’s Atlas of Protein Sequence and Structure is 50 years old. It has all the sequences printed together 😉 and some lovely structures, the definition of mutation data matrices and sequence databases. ⁦@UoDLifeSciences⁩ ⁦@emblebi⁩ ⁦@ISBSIB⁩

[Bill Pearson @FASTABill]

@suburban_queer West Provo?

[Bill Pearson @FASTABill]

@strnr These kinds of analyses can be very challenging. Many widely used software packages may not be available on GitHub, and "accuracy" can be very domain specific. Was BLAST one of the packages examined? How did it score on the "accuracy" scale? On the "development" scale?

[Bill Pearson @FASTABill]

@Real_Poobah @aaronquinlan Of course, all University presidents were originally Professors.

[Oded Rechavi @OdedRechavi]

Believing that the protein that you study is involved in every biological process

[Bill Pearson @FASTABill]

@DrSalinasLab I'm very surprised to read this. Most of the researchers I know are happy to give talks to students -- our students often recruit more famous speakers than our faculty. The problem is usually time commitments (though perhaps less so this year). No more than one trip per ???.

[Bill Pearson @FASTABill]

@peevedccprof I'm sure the idea is correct (we don't like paying teachers), but a quick look at the Mississippi Dept. of Education salary scales suggests it is not accurate.

[Sam Nicholls @samstudio8.bsky.social @samstudio8]

think about this the next time you get to decide to fund some bioinformatics research or not