1. Assembled genes: consensus gene sequences for each successfully assembled transcript that passed quality control filters.

2. Orthologous alignment: multi-species alignments of assembled transcripts, restricted to the regions of the genes present in all species.

3. Orthologous alignment coding regions: Similar to #2, but restricted to the coding regions only, and respecting the reading frame (as described in Supplemental Methods).

4. dNdS PAML output: For all genes and lineages meeting the criteria described in Supplemental Methods. Branch (lineage) field is as per depicted in the associated key (PDF file).

5. Gene expression data: includes normalized gene expression data, and expression likelihood values. Normalized expression values are provided for each individual. "NA" indicates that the gene was not assembled for that species. Expression likelihood values are output from the evolutionary model described in the text, for each lineage, for all genes that were assembled in a minimum of six species. NA indicates terminal branches for which that gene was not assembled in that species. Likelihood values were not estimated for the non-primate mammals, but their expression estimates were used in the evolutionary model and to determine whether a particular lineage could be analyzed (i.e., requiring at least two outgroup datapoints - see Supplemental Methods). For these species in the file, 0 indicates no assembly for that gene. 1 or 2 indicates assembly and expression data available for that gene. Likelihood values header names are explained as follows (in parentheses where not self-evident). human	chimp	hominid (human/ chimpanzee ancestral lineage)   	macaque	vervet	cercopith (macaque/ vervet ancestral lineage)	catarrhine (human/chimp/macaque/vervet ancestral lineage)	marmoset	anthropoid (catarrhine/ marmost ancestral lineage)	slowloris	mongoose	crowned	eulemur (mongoose lemur/ crowned lemur ancestral lineage)	BWlemur	eulemur_varecia (mongoose/ crowned/ black and white ruffed lemur ancestral lineage)	sifaka	eul_var_prop (mongoose/ crowned/ BW ruffed lemur/ Coquerel's sifaka ancestral lineage)	ayeaye	lemur (all lemur ancestral lineage)	strepsirrhine (lemur/ slow loris ancestral lineage)	primate (all primates ancestral lineage).

6. Individual SNPs. All identified SNPs for each individual and gene, considering all sites with sufficient coverage for SNP identification in that individual (see Methods and Supplemental Methods). The files are identified with the species and flowcell and lane information (see Supplemental Table 8 for a key to translate to other sample identifiers). Each identified SNP is recorded with the following information: Gene, position (corresponding to the position in the consensus assembly sequence for that gene, see files in #1 above), consensus sequence at that position, total numbers of reads, codon position (1, 2, 3 are in gene coding regions, 4 = UTR or in coding region but unknown [e.g., near indel], 5 is unknown [e.g., could not identify coding frame], read counts with nucletide A forward strand, C forward, G forward, T forward, A reverse, C reverse, G reverse, T reverse, the Alternate allele (of SNP call, versus consensus sequence).

7. SNP genotypes. SNP genotypes for each species, for all genes and sites with sufficient coverage for SNP identification in ALL analyzed individuals for that species. Fields are a gene/position unique identifier, gene, position (corresponding to the position in the consensus assembly sequence for that gene, see files in #1 above), consensus sequence at that position, codon position (see above), number of reads mapped to the site in total, and then the genotypes for all indivduals (n = 2, 3 or 4 depending on the species). 
