- Jayne's investigations revealed that blood data is whole blood, not PBMC
- large proportion of myleloid cells is expected because of this
- Many cells may in fact be neutrophils (large majority in the milk, many in blood)
- Because of the above discoveries, the weird cell proportions are not due to poor quality data and low UMI/gene counts are expected for these cell types.
- We can relax the QC filtering cutoffs to include more cells.
- Jayne added new marker genes and rough cell identities.
- Jayne started work collecting data sets to map our data onto. This will help with cell type assignments.
- One milk dataset of interest uses non ensembl geneIDs. Need to figure out which database these IDs are from and map to ensembl.
- Re-run QC with Jayne's relaxed cutoffs
- Use Jayne's markers to assign cell types and make a nice looking report.
- Figure out weird gene IDs in human milk data and map to ensembl.
from ensembl.org/Bos_taurus/Info/Annotation
Assembly The ARS-UCD1.2 assembly was submitted by Usda Ars on April 2018. The assembly is on chromosome level, consisting of 2,597 contigs assembled into 2,211 scaffolds. From these sequences, 30 chromosomes have been built. The N50 size is the length such that 50% of the assembled genome lies in blocks of the N50 size or longer. The N50 length for the contigs is 25,896,116 while the scaffold N50 is 103,308,737.
From: https://www.ebi.ac.uk/ena/browser/view/GCA_002263795.2 Assembly: GCA_002263795.2 This is a de novo assembly of Bos taurus using long reads for assembly and short reads for scaffolding and polishing. The assembly was scaffolded using a combination of in vitro chromosome conformation capture sequencing, optical maps, and a recombination map. Subsequent scaffolding was confirmed using genetic linkage maps and a Radiation Hybrid map. This material is based on work supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, National Research Support Project NRSP8 (Cattle Genome Coordination) Comment In May 2022 254 sequences were suppressed because they were found to be contaminants.
Organism: Bos taurus Accession:GCA_002263795 Assembly Name:ARS-UCD1.2 Assembly Level:chromosome Genome Representation:full Total Length:2,711,209,833 Ungapped Length:2,711,181,671 N50:103,308,737 Spanned Gaps:386 Unspanned Gaps:0 Scaffold Count:1,957 Count Contig:2,343 Contig N50:25,896,116 Contig L50:32 Contig N75:14,624,078 Contig N90:4,439,585 Scaf L50:12 Scaf N75:69,862,954 Scaf N90:51,992,305 Replicon Count:31 Count Non Chromosome Replicon:1 Count Alt Loci Units:0 Count Regions:0 Count Patches:0 ENA-LAST-UPDATED:2022-05-20
Gene annotation The gene annotation process was carried out using a combination of protein-to-genome alignments, annotation mapping from a suitable reference species and RNA-seq alignments (where RNA-seq data with appropriate meta data were publicly available). For each candidate gene region, a selection process was applied to choose the most appropriate set of transcripts based on evolutionary distance, experimental evidence for the source data and quality of the alignments. Small ncRNAs were obtained using a combination of BLAST and Infernal/RNAfold. Pseudogenes were calculated by looking at genes with a large percentage of non-biological introns (introns of <10bp), where the gene was covered in repeats, or where the gene was single exon and evidence of a functional multi-exon paralog was found elsewhere in the genome. lincRNAs were generated via RNA-seq data where no evidence of protein homology or protein domains could be found in the transcript.
Statistics Summary Assembly ARS-UCD1.2, INSDC Assembly GCA_002263795.2, Apr 2018 Base Pairs 2,715,853,792 Golden Path Length 2,715,853,792 Annotation provider Ensembl Annotation method Full genebuild Genebuild started Sep 2018 Genebuild released Dec 2018 Genebuild last updated/patched Nov 2018 Database version 107.12 Gene counts Coding genes 21,880 Non coding genes 5,235 Small non coding genes 3,375 Long non coding genes 1,488 Misc non coding genes 372 Pseudogenes 492 Gene transcripts 43,984 Other Genscan gene predictions 46,441 Short Variants 98,844,739 Structural variants 18,942